methods-corr.txt

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=1 p:cover ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=2 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=3 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=4 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=5 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=6 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=7 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=8 ]]
Methods Of Reproducing
Research Materials

 

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=9 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=10 ]]
Methods Of Reproducing
Research Materials

A Survey made for the Joint Committee on Materials for Research
of the Social Science Research Council and the American
Council of Learned Societies.

by

ROBERT C. BINKLEY

Western Reserve University

This edition published by Edwards Brothers, Ann Arbor, by the
photo-lithographic process from author's own manuscript,
pica double space typescript. reduced one-half.

ANN ARBOR 1931

Copyright 1931
ROBERT C. BINKLEY
(Printed In U.S.A.)

EDWARDS BROTHERS, INC.
Lithoprinters and Publishers
ANN ARBOR. MICHIGAN

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=11 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=12 ]]
FOREWORD

The Joint Committee on Materials for Research
of the Social Science Research Council and the American
Council of Learned Societies was organized in 1930, and
consists at present of the following persons.

Solon J. Buck,
Western Pennsylvania Historical Society,
Chairman

Robert C. Binkley, Western Reserve University,
Secretary

Laurence V. Coleman
Norman S.B. Gras
Waldo G. Leland
H.M. Lydenberg
Arthur Quinn

At its second meeting, held in September 1930, the
Joint Committee instructed its secretary to make a study of
the various methods available for the reproduction of re-
search materials, their comparative costs and their respective
limitations. The survey has been carried on with the help-
ful cooperation of many firms and individuals, among whom
special mention should be made of Dr. Thomas Martin of the
Library of Congress, Mr. H.M. Lydenberg of the New York
Public Library, Mr. J.W. Edwards and Mr. Eugene Power of
Edwards Bros., Ann Arbor, Mr. Ernest Kletsch of the Mime-
O-Form, the National Process Company of New York, and the
Washington Planograph of Washington, D.C., Mr. E.G.
Smithyman of the Photostat Company, Mr. Willard D. Morgan
of the Leitz Company, Mr. B.B. Snowden of the Ansco
Company, Mr. E.H. Naylor of the Rag Paper Manufacturers
Association, Mr. H.C. Lambach of the filmograph Corporation
Mr. L.B. Kennedy of Warren, Ohio, Dr. Miller Reese Hutchison,
Admiral Bradley A. fiske, Mr. H.K. Baumgardner of the Mid-
West Rotaprint Corporation. Mr. J.F. Perkins of the J.B.
Savage Co. of Cleveland, The William Feather Co. of Cleve-
land, Dr. B.W. Scribner of the Bureau of Standards, Mr.
Charles G. Proffitt of Columbia University Press, and Mr.
H.G. Deane of Yale University Press. The Spencer Lens
Co. and Bausch and Lomb Co. have helped with optical and
projection problems and with loan of apparatus. Other
firms, as mentioned in the text, have lent their aid.
Professors Paterson and Tinker of the University of Minn-
esota have helped in the study of legibility. To these
and others who have given their help the thanks of the
Joint Committee are due.

Robert C. Binkley
November 28, 1931

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=13 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=14 p:i ]]
TABLE OF CONTENTS

I. Publishing and Library Costs. . . . 1

The Book. Average retail price of general and
scholarly books by country of publication.

Analysis of book costs. Low factory cost, high
editorial and promotion cost.
Library costs. Acquisition and cataloguing;
housing.

II. Methods of Reproducing Research Material. . . . 8

Book printing from raised type as a standard;
new methods combining photography and lithog-
raphy. Revolutionary prospects in photo-
compositing and filmslide photography.

Printing from raised type. . . . 9

Linotype and monotype. Cost of 50,000 word
job. Short and long run methods. Multigraph.
Economies due to low-wage operation.
Mime-O-Form (See sample in appendix)

Table of quotations on cost of printing
50,000 words, 100 pages, 6 x 9 inches, by
various processes. . . . 15

The "Offset" Methods. Photography & Lithography. . . . 17

Principle of Lithography. Principle of the
Offset press. Rapid change in technology.

1. The Size of the Press. Economies of
small press; quality of product of
large press.

2. The storage of metal sheets. Relation to
control of size of edition. Policy of
Fac-Simile Text Society.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=15 p:ii ]]
3. Method of putting the image on the
metal sheet. . . . 20

a. Direct Impression
b. Direct Transfer
c. Wet Plate
d. Photo-Process - Negative
e. Photo-Process - Positive
f. Offset - Deep
g. Paynetype
4. The sensitivity of the emulsion on the
metal sheet
5. The combination of photo-lithography
with etching
a. Aquatone
b. Depth-o-tone
c. Pantone
6. Dry Lithoqraphy
7. Photo-Composing Machines

The Mimeograph and its Relatives. . . . 30
Impossibility of enlarging or reducing
size of copy. Coxhead Press

Table of quotations. Book of 50,000 words,
100 pages 8½ x 11 inches, photo-lithography
& mimeograph compared. . . . 33

The Photostat and its Relatives. . . . 34

Photostat costs not elastic with size of
edition. Cost of materials. Cost of
labor and chemicals. Comparative cost
data from New York Public Library and
Cleveland Bell Telephone Company. Exper-
ience of Massachusetts Historical Society
in copying newspapers.

Photostat Equipment. . . . 39

Comparison of Photostat and Rectigraph
The Luminiophor Plate.

Collotype and Photograph. . . . 42

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=16 p:iii ]]

III. Comparison of Different Methods of
Book Manufacture. . . . 44

The five variables of book production.
state of the copy in hand, typescript
or print        in the produc.
presence of non textual matter, degree
of legibility required, extent of market

1. Reprinting material which is ready
for photography

Dividing line between photostat
and photo-lithography

Table of comparative costs of
Photostat and photo-lithography
in edition of less than one hundred. . . . 45

Possibility of shifting from
photostat to photo-lithograph in
projects for copying old newspapers.

Critical point in photographic
copying where size of edition per-
mits sale at normal book prices. . . . 50
Practice of Fac-Simile Text Society,
British Shakespeare Association

2. Typewritten books. . . . 52
Redistribution of labor between
author and publisher. Advantage
of typescript in reproducing
mathematical or foreign characters
charts or pictorial matter
Problem of the right hand margin.
Relations of type-size, legibility
and cost.

The cost of legibility. . . . 56
Cost-per-word and cost-per-page area.
Psychological problem of legibility.
Experiment by Tinker and Paterson for the
Joint Commitee.

Chart showing relations of type size
number of words to the square inch

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=17 p:iv ]]

cost of publishing 50,000 words at normal book
prices (1¢ per page), and legibility compared
with newsprint. . . . 59

Table of cost and legibility. . . . 60

Use of the typewriter in obtaining various
costs and legibilities. Elite and Pica type.
Single, double and special spacing. Areas of
surface covered by typewriter. Reduction by
photography, four pages in one, or from large
page to a small one. Length of line.

Table of styles of typescript prepared for
offset press, with cost of 50,000 words at
1¢ per page. . . . 65

Problem of legibility as defined for further
study. Standards of legibility

Optical aids to legibility. . . . 66

Limitations of the reading glass.

The fiskoscope. Its place in connection
filmslide photography rather than book pro-
duction.

Quality of paper. . . . 68
Btitish and American attempts at standardiza-
tion. Content vs. Performance. Comparative
table of specifications. Rag Paper Manufacturers
point of view.

Comparative costs of ordinary book and permanent
book papers. . . . 73
Relation of cost of paper to size of edition
and risk of non-sale.
Possibility of special printings for libraries.

Recommendation on book production. . . . 77
Organization of purchasing power.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=18 ]]

IV. filmslide Photography. . . . 81

Two uses of filmslide: as an intermediate
step in photostating, and as a complete
reproduction method.

Cost of film. Perforated and non-perforated
types. Importance of reduction ratio in
comparison with photostat.

The physical limits of reduction ratio --
Resolving power of film in lines per millimeter.
Practical tests of legibility of material.
Degrees of legibility, A,B,C,D,E.

Area of film per exposure in the various
filmslide cameras. . . . 86

Price comparisons of equipment. . . . 88

Description of apparatus. . . . 39
The Vi-Cam
The Q.R.S.
Cinescope-Photoscope
Ansco Universal Still-film Copying Camera.
Leica and Lemare equipment compared
with Ansco. Copy holding. Double
exposure. Loading. . . . 95
filmograph and Kennedy apparatus.
Business side of the use of this
equipment. . . . 98
Proposed filmograph. . . . 
Library service. Societe dee editions
sur films de bibliotheques de France

Comparison of efficiencies in reduction.
Table of Reduction ratios. . . . 103

Projection apparatus. . . . 104
Leica, Delineaacope, filmograph, Kennedy,
Heat. Screens.
Direct reading of filmslides without
projection. . . . 106
Choice of makes of equipment on the market.
Need for a finer camera to take non-per-
forated film, and for a better projector. . . . 107

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=19 p:vi ]]

costs of operation. . . . 109
Table of high and low costs. . . . 111

Conclusions on copying books and manuscripts...112

Conclusions on copying newspapers. . . . 112
Special characteristics of newspaper prob-
lem. Perishable paper. Large surfaces.
Completion of files.
Cost of New York Times current file as a
standard of measurement.
Table showing costs of one hundred pages of
New York Times, wood pulp paper, rag paper, binding,
filmslide, photo-lithographed, and price of
Paris Temps and London Times. . . . 115

Note Taking Equipment. . . . 118
Scholar s problem as distinguished from
library's problem. Use of photo-lithography
in preparing Early Modern English Dictionary.

Use of filmslide and of filmslide combined with
enlargement. Martin magnifying glass.
Need for cheap photostat for note-taking

Recommendations. . . . 121

1. Library installation of filmslide apparatus.
Scientific check of conclusions of this survey
on comparison of Leica and Lemare. Installa-
tion of preferred equipment to supplement
photostat and take load off of borrowing
and lending obligations.

2. filmslides and the newspaper problem. . . . 122
Need for improvement in apparatus
before organizing large scale projects.
Hence:
A. Study of durability of film
B. Study of resolving power & reduc-
tion ratio. Test of Hutchison proc-
ess. Improvement of photographic
apparatus.
C. Improvement of projecting apparatus

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=20 p:vii ]]
V-

3. Photostat for Note Taking. . . . 124

4. Organization of the book buying market
in connection with book production. . . . 124
Correlation with Mr. Marshall's survey of
publishing facilities and revolving
funds. Opposition between scholarly
and commercial interests in publishing in
respect of extent of market, character
of product, and assessment for promo-
tion costs. Instances of publishing
needs of scholars. Starting point
of organization; scholars who have
publishing needs, also controlling a
quota of library funds. Organized
purchasing power in any field effective
when one to three hundred institutions
are included. Analysis of economies
from organized purchasing power.

Types of material. Group I in which libra`
ry interest predominates. Group II in which
specialists" interest predominate

Action by Joint committoe. . . . 131

A. Publishing projects in search of funds. . . . 131

A procedure to determine in each case
whether self-financing is possible, by seek-
ing a sponsoring organization. separating
editorial from publishing costs, and
publishing limited edition for a list of
subscribers.

The need for lists of special markets, to be
compiled from the Joint Committee"s Survey
of Activities of American Agencies in the
Improvement of Materials for Research.

Proposals of Fac-Simile Text Society and Edwards
Bros

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=21 p:viii ]]

B. Funds in search of publishing projects. . . . 133

Principles in respect of which policies
of existing publication agencies should
be defined --

1. typescript publishing 3. limited
market. Action of Joint Committee to
help to organize market for publishing
agency which accepts these principles.

C. Library collaboration in purchasing. . . . 134
Connection between problem of reproducing
materials and problem of controlling
government document acquisition, as
presented by Mr. Gerould

D. Newspaper reproduction. An experimental
project in photo-lithographing instead of
photostating an old newspaper file. A
project to provide half-size durable paper
reproductions of newspapers which are now
being bound by many libraries.

E. Model enterprises to be sponsored by the
Joint Committee.

A sample organization of research and purchas~
ing interest; a sample project of publication.

F. Further study or legibility and typography. . . . . 137

Distribution of This Survey. . . . 138

APPENDIX

Sample Mime-O-Form page
Sample of newspaper reduced one half

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=22 p:1 ]]

A SURVEY OF METHODS OF REPRODUCING RESEARCH MATERIALS

I. Publishing and Library Costs

The Book

The point of reference from which the
problem of reproducing research materials must be
examined is the system by which publishers produce
books and libraries acquire, arrange and store them
for use. The financial policies of institutions,
the stack room and reading room equipment of libraries,
the training of library personnel as well as the
research habits of scholars are standardized in favor of
the use of books of several hundred pages, from six to
eleven inches in height, costing the purchaser from one-
half to two cents a page.

There are other forms in which research
material is made accessible, but in any of these other
forms there is created a special problem of finance or
administration. Maps, newspapers, "oversize" books,
manuscript collections, films, collections of pamphlets
and broadsides present special difficulties and cause.
special expenses regardless of their acquisition costs.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=23 p:2 ]]

The normal acquisition costs of books
or bound volumes of periodicals which can be
shelved like books is something less than five
dollars a volume. In a public library which
purchases a large amount or popular fiction it
falls to $1.30 a volume. The list prices of books
per hundred pages, as given in standard annual
boot-lists are as follows:

AVERAGE RETAIL PRICE OF BOOKS, PER ONE HUNDRED PAGES
General Books Scholarly Books (History)

U.S. $.53 $1.20
Gt. Britain .86 1.20
Germany .75 1.04
France .58  .34 (unbound)
Italy .65

Research materials that can be presented in book form
within this price range can be fed into the library
system of the country without requiring significant
changes in present acquisition policies.

Analysis of Book cost

The cost accountancy of book production is
very complicated, but the essential fact is simple:
the unit cost of producing a book falls rapidly as the
size of the edition increases. This is the crux of the
problem that must be faced in any plan which will seek

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=24 p:3 ]]

to follow the requirements of research rather than
the demands of the market in deciding what materials
are to be reduced to book form. When the prospective
market and hence the size of the salable edition is
small, the unit cost of the volumes rises above the
level at which normal purchasing policies by institutions or 
individuals will float the enterprise or accomplish
adequate distribution. The critical test of book
cost, for the purposes at this survey, is the
minimum size of an edition which can be sold within
the normal price range.

Another striking characteristic of book
costs is the large percentage of the sales price
which goes to other objects than actual printing. A
great university press calculates that two fifths of
the cost of a book are strictly manufacturing costs,
one fifth is editorial and general overhead, and two
fifths selling costs. Another university press bases
its price calculation on the following procedure:

1. Take the "shop cost" of the book (prose-work,
paper, binding, etc.) A maximum and minimum
quotation for an edition of 1000, book or 100
pages, 50,000 words is . . . $370 $775

2. Add 20% for editorial and overhead costs, which
brings the amount to . . . $444 $930

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=25 p:4 ]]

3. subtract from the edition 100 copies for
review and promotion, and divide the total
cost by the salable edition (900 copies in
this case) to get unit cost of each book . . $ .49 $1.03

4. The sales price should then be fixed at
three times this unit cost of the book, if
the book is to return cost of publication
and a small royalty to the author ......... $1.47 $3.12

The expenses of distribution are between 50%
and 33% of the gross receipts.

The commercial publishers devote less
attention to scholarly values, but they find their
costs distributed in about the same way. The
following calculations, from two independent
sources, are substantially in agreement:

Cost of book sold at $2.50 $2.00

Manufacturing cost ........ .45 .......... .40
Author's royalty .......... .25 .......... .20
Editorial and promotion ... .51 1/2 ....... .40
Profit (Publisher) ........ .28 1/2 ....... .20
Retailer's discount ....... 1.00 ....... .80

Publishing at this cost level calls for the sale of
editions of two thousand.

The Carnegie Endowment for International
Peace, which distributes much of its material free of
charge, and does not therefore figure promotion as

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=26 p:5 ]]

part or its costs, allows $3800 to pay for an
edition of 1500 copies or a cloth bound book of
596 pages in the International Law series. This
is at the rate of $.42 per 100 pages, or almost
exactly one third of the average list price of
American published books of the same class. These
figures contribute to the rough generalization that
book prices as listed are made up of three equal
parts:

One third to write, print, bind and edit the book.
One third to move it from publisher to retailer.
One third to move it from retailer to purchaser.

Libraries often get discounts which reduce the last of
these price increments.

These figures indicate that economies in the
printing process, unless accompanied by economies in
the system of distribution, will not greatly alter the
present limitations which impose themselves upon the
publishing of research materials. If these distributing
costs could be eliminated entirely, the critical size of
the minimum paying edition would drop from around two
thousand to around three or four hundred, without any
change in the production method.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=27 p:6 ]]

Library costs.

The purchase price of the book is not the
heaviest expense that has to be met in making it
available to a scholar in a library. Library
finances must be given a place in the total picture,
because different methods of reproducing materials
fit differently into the scheme of library finance.
But the cost analysis of library administration is a
disputed field, in which the files of the Library
Journal report a prolonged controversy. The most
active work in cost analysis has been applied to the
study of cataloguing cost -- the cost of putting the
book on the shelves. The figures which seem best
attested are those of the University of California
(72¢) per book. The general administrative expense
of a library cannot be pro-rated with any accuracy
among individual books, but must be taken into
account. figures collected by the Government show
that university libraries spend from 50% to 72% of
their funds for other than acquisition purposes
(50% Cornell, 72% Yale). The normal amount is about
two thirds. But the capital cost of building and land
is not included in this figure. A library like the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=28 p:7 ]]

New York Public Library, which must take into
account a high cost of land, maintains an enormous
capital investment in land and buildings. All
libraries must be expanded from time to time to keep
pace with the increasing quantity of material and
numbers of readers. If the financing of this expansion
(which may amount to the doubling of the plant every
twenty years) be calculated as an element of library
expense the capital cost may be as high as double the
administrative coat, or four times the cost of purchas-
ing books. These considerations become especially
important in connection with manuscript or pamphlet
material, for which the administrative cost is high, --
or in connection with bound newspaper files, against
which an abnormal housing charge must be assessed. It
can be roughly calculated that the expenses of bringing a
book from the press to the scholar's cubicle in the
library multiply at such a rate that for every hundred
dollars devoted to library facilities for study, the
amount that goes into book production may be anything
from a minimum of 3.50 to a maximum of 14.00. One
dollar spent to manufacture a book (royalties, press -
work, binding, and editing) must have from six to

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=29 p:8 ]]

twenty-eight dollars added to it on account of library
costs.

II. Methods of Reproducing Material

Just as the policies of publishers and
librarians set business standards for all enterprises
in the field of scholarly publication, so printing
from raised type sets a technological standard. But
printing from raised type is now being challenged by
other methods of reproducing texts. The new methods
combine in various ways the principles of lithography,
typewriting and photography, or make use of photog-
raphy alone. Substitutes for printing are so many
that the whole technology is in a state of flux, with
prospects of revolutionary change. Two revolutionary
processes, on the borderline of practical application
today, are photo-compositing (preparation of lithographic
master-copy by photographing letters directly on a
page), and filmslide photography (reduction of text to
a miniature slide, to be read by projection on a screen).
The inventors and promoters of the new devices are
concerned very little with the scholar's problem.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=30 p:9 ]]

Lithographic printing methods are commercialized in
response to the needs of advertisers, and filmslide
photography is receiving its most successful develop-
ment in connection with the needs of abstracting and
title-guaranty companies. If the scholar's needs are
met by the new processes, it is only in an incidental
or accidental way.

Printing from Raised Type

Typesetting for ordinary straight printing
of text can be done most economically on one of the
composing machines, which are of two kinds. The
most common cast type in slugs of one line. (Mergen-
thaler, Intertype, Ludlow (for display work),
Typograph, and Victorline). The Discotype, which is
not yet on the market, is a machine for casting
separate type pieces of 1/2 depth and then uniting
them in a slug. The Monotype, together with its
defunct relative, the Stringertype, casts the letters
separately but in correct order, so that it can be
used either as a type foundry or for setting straight
copy. Hand-set type is used principally in connec-
tion with advertising compositions. It is a luxury

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=31 p:10 ]]

when used in straight textual matter.

The estimating of printing costs is a very com-
plicated process which has been standardized by the Unit-
ed Typothetae of America. The typographic cost follows
the number or words rather than the size of page or of the
type. A 50,000 word job of linotyping estimated at $5.00
per machine hour will cost $170 set in 4½ point type, and
$180 set in 7 point (newsprint) type. (See Fred W. Hoch:
"The Standard Book on Estimating for Printers", United
Typothetae of America, Washington, D.C.)

For any ordinary short run, the type as set in
its flat form is used. But if the greater running efficien-
cy of the rotary press is to be obtained, the type face
must be transferred to another printing surface, either by
making a matrix for casting a stereotype or by making an
impression in wax, which is then coated electrolytically
with copper to make an electrotype. Another device for
transferring from the type as set to a new surface is
to print it on a thin metal sheet for use in the offset
press. This third method will be discussed in the next
section.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=32 p:11 ]]

The interest of scholarship in cheap and
small editions is most nearly approximated in the
commercial world by the demand for a small office
printing equipment, of which the Multigraph is
the standard for raised type printing, and the
Rotaprint for offset printing.

The merit of the Multigraph is that
it can make printing look like typewriting through
the use of a ribbon, which is an achievement
diametrically opposed to the wish of the schol-
ar, who would like to see typewriting look like
printing. Multigraph typesetting is usually done
by hand, but with the "Compotype" apparatus
characters are embossed on strips of aluminum and
the strips assembled on a flexible sheet metal blank-
et which is then printed on the Multigraph. Th»
economies in Multigraph over print shop work are to
be found in the wage scale of the operators. Quo-
tations from a multigrapn company upon straight text
work are equal to the highest printer's quotation.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=33 p:12 ]]

Mime-O-Form.

There is a modification of the Multi-
graph developed by Mr. Ernest Kletsch in Washing-
ton, D.C. and called the Mime-O-Form. A Monotype
typecasting machine has been equipped to cast
type for a Multigraph. The Multigraph printing
method is so economical for short runs that its
prices compare favorably with the cheapest print-
ing or offset printing quotations and are not much
higher than mimeographing. Its product is a book
in ordinary type face, almost indistinguishable from
the product of a regular book printing press. (See
sample page in appendix)

So great is the variation in prices
quoted for printing small editions that it is evident
that the business and labor policies have as much effect

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=34 p:13 ]]

as the technical process itself upon small edition
costs.

The following tabulation of sixteen printing
quotations illustrates the variation. The Feather
Co. and Savage Co. are Cleveland firms doing a book
printing business; the Polk Co. is a Multigraph firm;
the Mime-0-Form is the combination of monotype and
multigraph referred to above; J. Gamber is a publisher
in Paris, France who specializes in university thesis
work and hence in small editions. His prices are
here corrected to allow the 10% addition for English
language work and the 25% customs duty. Edwards
Brothers, the National Process Co., Washington Plano-
graph, and Coxhead Corporation are users of the offset
method, which will be further described in the next
chapter. It must be explained at this point, however,
that offset printing quotations are of two kinds,
representing two classes or work. The higher price
is asked for the reproduction of a text which the
offset printer copies on a typewriter to make a
master-copy from which he makes, photographically,
his printing surface. The lower price is for the
reproduction of a text which goes to the printer as a

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=35 p:14 ]]

suitable master-copy, either the author's clean type-
script or the pages of a printed book. The higher
quotation includes a typewriting charge. The
quotations from the Lund, Humphries (London) are taken
from G.B. Harrison's paper on "Facsimile Reprints", given
before the Anglo-American Conference of Historians,
13-18 July, 1931.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=36 p:15 ]]
[[table ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=37 p:16 ]]

The various processes thus take the following
order as regards cost. Printing in France is cheapest,
and offset printing by Edwards Brothers next. Then
comes Mime-O-Form in editions up to 1150, and low-
quotation printing in editions above that figure. The
size of an edition necessary to reduce book cost below
a dollar a hundred pages runs from 50 (French printing)
to 525 (Multigraph). It is obvious that many factors
other than technique influence these prices. The
"Ratios for Printing Management" gathered by the
United Typothetas of America show that only 73% of the
income of a printing business goes into production or
"factory costs". This 73% is made up of fixed charges
(6%), materials (34%) and running costs (33%). The
differences in technique would affect principally
certain subdivisions of this running cost item, which
is only 33% of the total sales price. In small edition
printing the proportion applied to materials falls as
low as 20%. New technical processes often exert their
greatest influence on price through their use of
untrained or half trained labor.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=38 p:17 ]]

The 'offset'' Methods. Lithography and Photography.

The most interesting development in contemp-
orary printing technique is the encroachment of
"offset printing" upon printing from raised type.
Offset printing has developed in the last thirty
years from the combined application of principles of
lithography and photography, together with a new type
of printing press.

Lithography was developed by applying the
principle that grease repels water but attracts a
greasy ink. A stone surface was covered with the
greasy marks of a lithographic pencil, then mois-
tened, and then inked. The ink would adhere to the
pencil marks, avoid the moistened spaces, and transfer
to paper as a lithographic print.

This principle is now applied by substitut-
ing a thin sheet of zinc or aluminum for the
lithographic stone, a coating or a light-sensitive
albumin or bitumin compound for the lithographic
pencil, and photography for the craft of the
artist. The metal plate with its albumin or bitumen
coating acts exactly like photographic paper except
that it requires a much longer exposure to light.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=39 p:18 ]]

Printed or typewritten pages, together with illustra-
tions, are copied on the metal either directly from the
object or indirectly from a film negative. The metal
sheet then takes the place of the form of type used in
ordinary printing. But the printing method is diff-
erent for two reasons. The metal sheet must be mois-
tened after every impression and before every inking
because the process is fundamentally lithographic.
And the printing surface of the metal sheet is pro-
tected from abrasion by the "offset" method of
applying ink to paper:- the metal sheet prints on a
rubber blanket, which transfers the image to paper.
It is this last feature of the complicated new
technique which has given the whole process its name
of "offset printing", although there are certain
variant techniques which print direct without the
rubber blanket, and one device for omitting the
moistening between impressions.

New inventions and appliances in the field
of offset printing are patented every month. The
principal significant differences in equipment, and
the chief lines of technical development, are the
following:

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=40 p:19 ]]

1. The Size of the Press.

The smaller presses, such as the Coxhead or
Schwarz Press and the Rotaprint are better adapted for
small rum work than the larger presses such as the
Harris Press. The larger presses do better with long
runs and with color work. There are many types of
offset press, just as there are many types or flat-bed
press, and the same company will offer models in
different sizes. The large press is finer, the small cheaper.

Examples of the small presses are these run
by Edwards Brothers, Ann Arbor, Michigan, and Mid-West
Rotaprint Corporation, Cleveland, Ohio; examples of the
large press are those run by National Process Company,
New York City and by Washington Planograph co., Washing-
ton, D.C. All of these firms have collaborated in
this examination of their processes.

2. The Storage of the Metal Sheets.

Some presses, as the Coxhead or Harris presses,
use a rather heavy metal sheet which can be cleaned and
refinished after every printing. Such sheets provide
a saving in that it is cheaper to clean the old than
to buy a new sheet, but they affect the printing costs
because they make it expensive or impossible to store

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=41 p:20 ]]

the sheets for a fresh printing of a book which has
already been run in one edition. Rotaprint and
Edwards use a thin, comparatively inexpensive metal
sheet which can be stored, a hundred sheets in an
envelope, and used to bring out reprints. This fea-
ture of the process makes possible a greater elasticity
in fixing the size of an edition. The Facsimile Text
Society, which has its work done on the Harris Press,
prints editions of 1000 even if its sales prospects do
not call for that number, because once the printing has
been done, no reprints could be secured without making
fresh plates. This tends to cause resources to be tied
up in a large stock of finished books. There is a
similar project for the printing of small working
editions of rare old English books to be used in the
preparation of a dictionary of 16th century English.
In this project the Rotaprint type of press is used, and
the editions are limited to two hundred copies. But
since the metal sheets can be stored, further printings
can be made inexpensively, it needed.

3. Method of Putting the Image on the Metal Sheet

a. Direct Impression

This is the closest relative to the original

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=42 p:21 ]]

lithographic process. Pen and ink or a typewriter with
a special ribbon directly on the metal is used instead
of a lithographic pencil upon stone. The typewritten
letters accept ink and reject water; the rest of the
surface accepts water and rejects ink.

b. Direct Transfer

Instead of being typewritten or drawn directly
on the metal sheet, the image is made upon paper with a
greasy ink and then transferred to the sheet by moisture
and pressure. This is one of the methods used by the
Coxhead Press.

c. Wet Plate

The image transferred to the metal sheet is
first made by photography upon glass instead of ink on
paper. This is the finest method for copying difficult
objects such as badly faded newspaper pages or half tone
prints. The process which goes under the name of "Elka"
is a wet plate process.

d. Photo-Process Negative

Instead of removing an image from one surface
(glass or paper) to another, as in the direct transfer
or wet plate process, the metal sheet is given a coating
of a light sensitive substance, based on bitumen, fish

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=43 p:22 ]]

glue or egg albumin, and an image is imposed upon it with
light, as in photographic printing, from a negative or a
positive. The most common processes make use of a photo-
graphic negative, which "fixes" the letters in bitumin and
albumin upon the metal sheet. The negative may be either
film or paper. film is in most common use in America, but
a special "Kodalith" paper is now manufactured by Eastman
for use in the photo-process. The use of Kodalith paper
cuts the cost in half.

e. Photo-Process Positive

In the usual photo-process the light passes
through the transparent parts of a film or the white parts
of Kodalith paper to fix the albumin or bitumen, which will
hold the ink. In the "positive" process the metal sheet is
coated with a light-sensitive composition which rejects
ink. The photographic process thus fixes an ink-rejecting
covering on the space between the letters, and leaves the
metal free to receive the ink. When the metal is inked
the special surfacing on the "white" part of the sheet is
dissolved away, the sheet is moistened, and thus prepared
for lithographic use, having moisture in the space which
is to print white, and ink in the space which is to print
black. In using this positive process it is sometimes
desirable to grease the paper from which the image is

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=44 p:23 ]]

being copied in order to render its white parts more
transparent. The Coxhead Press uses this positive photo-
process.

f. Offset-Deep

A secret preparation is on the market which
can be used as an emulsion in making a positive print,
and which will resist the action of acid. It thus
becomes possible to etch the metal which is to carry the
ink. This process results in deeper colors and blacker
print. It resembles the Depth-o-tone and Aquatone methods
which are discussed below as combinations of etching with
photo-lithographic technique.

g. Paynetype.

A device which photographs directly upon the zinc
plate, which is held in the camera. This is a variety of
the positive photo-process which is used in art work. It
resembles Collotype, a process which will be discussed in
connection with photography.

The most important difference between these
methods is that which affects the possibility of enlarging
or reducing the copy in the course of the reproduction
process. Direct impression and direct transfer methods
do not permit any alteration in size as between the original
and the copy. The Wet Plate Method and all the negative

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=45 p:24 ]]

photo-process methods permit of enlargement or reduction of
the copy. The positive photo-process method in most cases
loses all its advantages over the negative method if it is
necessary to enlarge or reduce the copy.

4. The Sensitivity of the Emulsion on the Metal Sheet.

With a few exceptions all kinds of photography
upon the albumin or bitumen compound on the metal sheet
required exposures of 15 or 20 minutes when the negative is
in contact with the metal sheet (the usual method) and 30 to 45
minutes when the image is transferred by projection. An
intensive search for more rapid emulsions has gone on for years.
Certain firms now have secret processes for speeding up the
emulsion. The Copyfire Co. of Detroit is said to have cut the
exposure time to three minutes, and Edwards Brothers, with the
help of their research department, have cut the time to one
minute twenty seconds. These more highly sensitized solutions
are not yet standard for use in the industry. The salesmen
for the Rotaprint press and process still tell their customers
that the time required is fifteen to twenty minutes.

5. The Combination of Photo-Lithography with Etching.

The object of development along this line is three-
fold: to get better art work, to make a more dur-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=46 p:25 ]]

able metal sheet which will stand a longer run in an
offset press or resist the abrasion resulting from
direct application to the paper in a rotary press, and
to touch up defective metal sheets.

a. Aquatone is a process for increasing the water-
absorption of the wet parts of the sheet, to produce a
clearer white in the print. The same principle, without
the name, is used in clearing spots or blurs in ordinary
photo-lithographic printing.

b. Depth-o-tone is a trade name for a process for
deepening the ink carrying parts of the plate and then
filling them with ink-absorbing substance; in art work
it produces a deeper color, and in ordinary printing it
"brings out" lines in the print which would otherwise
fail to register.

0. Pantone is a complicated process which is designed
to make use of a principle similar to that of photo-
lithography to prepare a printing surface which will
stand long runs and be very resistant to wear. It
makes use of the fact that chromium is a hard metal
which will accept and transfer ink, while mercury will
refuse ink. The chromium serves for the grease of

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=47 p:26 ]]

primitive lithography, mad the mercury for the wet
surface of the stone. Although this process will not
be useful in cheapening small editions, it is interest-
ing as an illustration of the variety of techniques
which can be used in printing. A sheet of iron is
coated with a thin surface of copper, and a very thin
surface of chromium on top of the copper, and then a
photographic emulsion on top of the chromium. The
text is transferred to the plate as in the usual photo-
lithographic processes, and the parts of the metal which
ordinarily receive water are etched with acid through
the chromium to the copper. Silver is then deposited
on the plate, and the silver adheres to the copper but
avoids the chromium. Then mercury is rubbed over the
plate, and the mercury amalgamates with the silver.
The result is an all-metal plate which will print
without using the raised type principle.

6. Dry Lithography

An inventor by the name of Price hes devel-
oped an ink which carries in itself the moisture
necessary to keep the wet part of the metal sheet from

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=48 p:27 ]]

taking ink. The use of this ink simplifies the printing
process by leaving out one step, namely, the moistening
of the metal between every impression. There is some
similarity between this and the Pantone process, because
in Pantone a little mercury is mixed with the ink to
keep the mercury surface replenished.

7. Photo-Composing Machine:

The offset press is sometimes used, in large
editions, for printing compositions which have been
made with raised type, and transferred to the metal
sheet, either for the sake of more efficient pictorial
effects, or to avoid the cost of "makeready" -- that is to
say, the wedging up of a printing face to secure an even
impression in a flat bed press. In this large-edition
use the photo-lithographic metal sheet is a rival of
the stereotype and the electrotype. As at present
organized this shop practice does not enter into the
calculations of small-edition manufacture. But it
illustrates the fact that raised type can be completely
dispensed with in the printing industry if any other
method than typesetting can be found for making the
master-copy to be used in a photo-lithographic process.

There are three kinds of work in which master

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=49 p:28 ]]

copy is now furnished without typesetting. Advertising
display work is sometimes prepared by the artist who does
his lettering just as he draws his pictures, and has no
use for type. Circulars, tabular matter, form letters,
and trial editions of text books are prepared as master
copy on the typewriter. The limit to the application of
this method to book printing is set by the unwillingness
of the reader to tolerate the uneven right hand margin.
The third kind of master copy is the printed page which
is being reprinted by the photo-lithographic process.

There are two approaches to the problem of
making an all-purpose master copy without using raised
type. One is to improve the typewriter by some method
of justifying right hand margins. The other is to
avoid the use of ink entirely, and photograph letters
in sequence upon a negative which can then be transferred
directly to the metal sheet.

A modified typewriter has been developed by
the Typary and Typon Company, London. This "Typary"
machine sets a line of type which is then automatically
spaced to justify the margin, inked, and printed on a
strip of paper, from which it can he transferred to the
photo-lithographic metal sheet. This machine has been

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=50 p:29 ]]

used in book manufacture. A similar invention. which
has not yet been given industrial promotion, is
Walton's type composing machine, which writes directly
on the metal sheets for offset printing.

The inventions which seek to "print with light"
are numerous. The Dutton Photoline sets letters which a
camera photographs line by line; the Bawtree apparatus
photographs a text on bromide paper for transfer to the
metal. Mr. G.F. Bagge of Cleveland has a similar inven-
tion under way. The two machines which seem to be
nearest the point of practical use are the August-Hunter
Photocomposing Machine which is being developed in
England, and the "Luminotype" invented by a Hungarian
technician by the name of Edmund Uher and manufactured
by the Maschinenfabrik-Augsburg-Nuernburg. One model
of the Uher machine had been constructed when the
recent financial crisis in Central Europe interrupted
the venture. The Uher machine photographs a line of
letters on a narrow strip, and then projects from this
strip, line by line, on a photographic negative. A
cheap substitute for film has been invented for the
purpose. The success of these photo-typing machines
would constitute the most decisive revolution in

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=51 p:30 ]]

printing since Caxton's time.

The Mimeograph and Its Relatives.

The techniques of reproducing copy vary in an
almost continuous sequence from the great planograph presses
which use highly skilled labor to produce long runs of fine
color and art work to the humble mimeograph with its wax
stencil prepared by a girl who has had a six months' course
in typewriting and its short runs of smudgy copy run off
by an office boy who has learned to operate the machine in
half an hour.

From the technical standpoint the distinctive
feature is the use of a stencil cut by means of a typewriter.
The sharpness of the impression can be increased by
improving the composition of the stencil. A.B. Dick, Chicago,
and the Gestetner Duplicator, Bloomfield, N.J., both offer
an improved stencil with a surface which "presents much
the appearance of a thin sheet of white rubber."

From the standpoint of performance the distinctive
feature of mimeograph technique is not the stencil method
but the limitation that the copies are turned out in the
same size as the original. From this standpoint there is
very little difference between the best mimeograph work

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=52 p:31 ]]

and the work done with the offset press from metal
sheets which have been typewritten with a special ribbon,
or the work done by the Coxhead Press by transferring
the text direct to the metal sheet from a typescript
made with a transfer ink, or even the alternative semi-
photographic method used in the Coxhead Press, which is
described as follows:

The printing is done by means of a specially
prepared zinc plate on which can be transferred any
kind of oririnal copy, old or new, hand or type-
written...a drawing of any sort. The original is
mounted on a zinc plate which has been sensitized
with a special solution. The plate is then exposed
to the rays of an electric arc lamp for a few
minutes. After this exposure the original is taken
off, undamaged by the process...After the plate is
developed and fixed, it is etched, rubbed with a
special ink and finally coated with a rubber
solution.

The products of these different methods range above the
mimeograph in quality of product according to the
following scale: the mimeograph with common wax stencils
is lowest, the improved stencils next, the product of
the Coxhead Press is next in quality, and finally the
offset printing from a typescript made directly on the
metal sheet. But as it will appear when the application
of these mimeograph methods to the scholars' problem
is analyzed, it is very inportant that none of then

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=53 p:32 ]]

change the size of the copy by enlargement or reduction.
For the use of photography in these various techniques
has two kinds of value. In the Coxhead process it copies
figures, signs, diagrams, and other matter which the
typewriter will not cut in a stencil. In the complete
photo-lithographic process it accomplishes not only
this service, hut also the enlargement or reduction of the
page as a whole.

The members of the mimeograph family have
received most of their employment as office appliances,
where their costs of operation have been dependent upon
wage scales and continuity of operation. The following
quotation is the lowest that has been received for
mimeograph service offered commercially by an outside
plant. The quotation is here given for purposes of
comparison with the lowest quotations for other comparable
kinds of work.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=54 p:33 ]]

Unit Cost of a Book of 100 Pages, 50,000 Words,
on Full Typewriter Size Paper (8½ x 11 inches)

Number of copies in edition: 50 75 100 200 300 400 500 1000

Photo-lithography from the
author's master-copy. (Price
not affected by number of
words on the page).......... $1.65 .98 .75 .64 .57 .44

Lithographic offset press
without photography 500
words typewritten direct
on the page........... 1.50 .90 .70 .60 .54 .42

Mimeograph on bond paper... 1.30 .80 .63 .55 .50 .40

Mimeograph on plain paper... 1.20 .75 .60 .53 .48 .39


Reprints from lithographic
metal sheet (either photo-
graphic or typewritten)..... $1.40 1.04 .85 .58 .48 .44 .41 .35

Reprints from mimeograph
stencils.......... 1.00 .80 .70 .50 .43 .40 .38 .34


[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=55 p:34 ]]

In comparing this table with the table of prices of
offset and raised type printing it must be noted that
this is a larger page. The larger area of paper for
the same number of words results in an increase of
legibility, or an increased space for diagrams or
tabulations.

Some use has been made of the mineograph
stencil for duplicating catalogue cards. The Elliott
Addressograph can be used for this purpose. A very
cheap and serviceable device of this kind was invented
by Hr. Hodgson when he was librarian of the International
Institute of Agriculture in Rome.

The Photostat and Its Relatives.

The peculiarity of the photostat when it is
compared with other methods of making copies of books is
that the unit cost does not diminish as the size of the
edition increases. It is made up almost exclusively of cost
of materials and running costs. The only special expense
which may be distributed over a number of copies is the
cost of a white-on-black negative in those cases where the
purchasers want black-on-white positives, or where there
is a saving of labor in "imposing" the negative. If one

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=56 p:35 ]]

of the purchasers is willing to accept the negative
copy, its cost need not always be "distributed" over the
positives.

The cost of materials in photostat work is the
determining factor, for if that cost were greatly reduced,
labor saving appliances would probably come into use which
would permit of an equivalent cutting of labor costs.
There is some discussion of a possible new photostat
paper, which will use a mercury compound rather than a
silver compound in the emulsion, and sell for such a low
cost that it  compete with engraving in book illustra-
tion work. At present one must reckon with the various
kinds of photostat paper sold by the Photostat Corporation
and its competitors.

The price of ordinary photostat paper per square
foot is $.042. But in the preservation of research materials
use should be made of the special "H" or "HH" grade of
permanent paper stock, manufactured by Byron Weston Company.
The "H" grade which is coated on one side sells for $.052
per square foot; the "HH" grade, which is coated on both
sides, and hence will bring a saving of half the cost of
binding the volumes, sells for $.10 per square foot of
paper, or $.05 per square foot of sensitive surface.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=57 p:36 ]]

When wastage has been taken into account, and
the variation due to the different ways of imposing the
negative and cutting the paper from rolls of different
widths, has been reckoned with, the cost of materials
for photostating a 100 page book in different page
sizes, and with different grades of paper, is as follows:

Dimensions 
of book
page in
inches

Cost of material in dollars

Ordinary stock "H" stock "HH" stock


4-1/2X 6 .78 to .83 .95 to 1.04 1.05
5-1/2X 8-1/2 1.42 to 1.51 1.87 to 1.95 1.84 to 1.91
6 X 9 1.63 to 1.83 2.03 to 2.06 2.34 to 2.35
8-1/2X 11 2.84 to 3.00 3.66 to 3.73 3.64 to 3.81
(typewriter
size)

11 X 14 4.67 to 5.01 5.80 to 6.22 5.91 to 6.06
11 X 17 6.02 to 7.68 7.45 to 9.55 7.29
13 X 22 12.01 14.90
(page of
N.Y.Times)

To these costs of paper there must be added
the costs of labor and chemicals. The New York Public
Library end the printing department of the Cleveland
Bell Telephone Company have very kindly contributed

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=58 p.:37 ]]

statements of their operating costs. In the Public
Library the great majority of the prints were 11½
x 14 inches, usually one copy only, and the average
cost of the paper was $.055. In the Telephone Company
the average size of paper was 8½ x 11, and in many
cases several copies of the same object are required.
The cost of paper averaged $.03 per page. In the
Public Library the $.055 worth of paper requires so
much reference and clerical work, that the cost of
the finished prints rises to $.197 each. In the
Telephone Company, where only operating room charges
need be considered, the $.03 worth of paper is worked
up into a print which costs $.063. This means that
the labor, depreciation and chemical supplies cost
$.033. These operating room costs in the Public
Library amount to $.06. The figures in total
operating cost are as follows:

Library Telephone Co.
Paper (per 100 pages) $5.50 $3.00
Operating costs (do.) 6.00 3.30
$11.50 $6.30

Cost per square foot
for Paper, Labor &
Chemicals $ 0.10 $0.084

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=59 p:38 ]]

The Public Library spends an additional $.082 on every
page to cover clerical help, postage, and desk supplies --
charges which are due to the special character of the
Library work. The unescapable operating costs are 10%
more than the cost of the paper, and may be more than
twice the cost of the paper. This percentage is
reduced as the work becomes more continuous. The Life
Insurance Companies operate with highly standardized
copy; their labor costs should be very low, but the
writer has not bean able to secure statements from
them. The work of the Massachusetts Historical Society
in copying colonial newspapers made use of the photo-
stat at its highest efficiency for the reproduction
of research materials. A letter from the Director, Mr.
Worthington C. Ford, explains the financing of this
enterprise:

'...we gave our operator the highest photostat
salary paid within my knowledge and with that
the cost of material represented twice the
cost of labor. I roughly estimated that it
cost us about 13¢ to make a page of a news-
paper of the large size...but we printed some
which required the largest size of sheet
adaptable to the photostat machine, and the
average cost was not increased. We sold the
newspapers at a fixed price of 15¢ a page.
Some years there was a profit of $300-$400 and
some years we came out even."

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=60 p:39 ]]

Similar newspaper photostating by Michigan University
Library and John Carter Brown Library are being carried
on at a rate higher than that charged by the Massachu-
setts Historical Society. The Library of Congress
estimates that its photostat copies of foreign archive
material cost it 15¢ each, and the British Museum and
Bodleian Libraries show a cost of 6¢ a page quarto size.
It can be concluded that every dollar's worth of photo-
stat paper requires from 50¢ to $1.10 in operating
expenses.

Photostat equipment

The two appliances which compete in the field
of photographic copying are the Photostat and the Recti-
graph. The Photostat is the older, and has succeeded
in making a common noun of its trade name. The differences
between the two kinds of apparatus do not seem to be
sufficiently important to have a decisive effect upon
costs, but are worth noting because they illustrate
certain of the practical aspects of photostat work.
Both firms have two kinds of apparatus, a stan-
dard equipment which leaves it to the operator to shift
the prints to a rinsing tank and to dry them, and a

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=61 p:40 ]]

"Continuous Photostat" or "Super-Rectigraph" which carries
the print automatically through the whole process from
exposure to drying. The highest priced continuous Photo-
stat costs $4,000. The claim is made that it will turn
out three prints a minute.

The following are the principal differences
between the two machines. The Photostat carries the
print on an endless chain through a developer bath, then
through a chromium roller which squeezes out the surplus
developer, into a rinsing tank of running water, and then
face-down into a shallow fixing tank which is continuously
agitated by a motor. The Rectigraph apparatus grasps the
end of the paper and pushes it down into the developer,
withdraws it up again by the same end, passes it through
a roller (but not a bath of running water), covers its
surface with the fixing solution and lets it emerge face
up in a deep fixing tank. All these features are inter-
related. The Rectigraph uses a deeper tank because the
solution is allowed to lie still; the Photostat agitates
the tank, which must therefore be shallow. The Recti-
graph claims that it can dispense with the agitating of
the fixing bath because it covers the whole surface of
the paper evenly with the solution before it enters the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=62 p:41 ]]

bath. It is this covering of the surface prior to
immersion in the bath which permits the print to appear
face up in the tank. The Photostat has a much heavier
frame, which is necessary to resist the vibration
caused by the agitating of the tank, but which has many
independent advantages. The Rectigraph wastes more
paper than the Photostat because it requires more of a
margin at the top of the sheet when dipping it into the
developer. Both companies claim that they are more
saving of chemicals; in the judgment of the writer the
Photostat has the better claim because the developer is
washed off in running water before the print enters the
fixing tank. Correspondence with two users who have
each had experience with both types of machine reveal a
preference for Photostat, partly on account of the more
extensive service organization.

A type of equipment related to the Photostat,
but making copies without camera or lens, is the
"Luminiophor" radio-active plate which sends out rays
that penetrate an ordinary piece of white paper as if it
were a transparent film, but which do not so readily
penetrate through an ink surface. The plate is used in
a dark room, by placing the master-copy on top of the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=63 p:42 ]]

sheet of photostat paper and the Luminiophor plate on
top of the master copy. The result is a contact print.
The difficulty arises naturally in the case of paper
which is printed on both sides. Even for such pages
a negative of the printing on one surface can be made
by reflection, after cancelling out the effect of the
print on the other surface by laying the page on black
paper. A true copy can then be made from the negative.
The Luminiophor plate must be recharged every few
minutes by exposure to sunlight or strong electric or
magnesium light. It sells for 25¢ per square inch.
Its place in the scholar's equipment could only be as
an aid to note taking, and even there it is inferior in
value to the Leica, Ansco, or Lemare equipment for
making filmslide copies of documents.

Collotype and Photograph.

The more costly but finer relative of the
photostat is the photograph, and the aristocratic cousin
of the photo-lithographic process is the collotype
process. In collotype work a film (of gelatine) is
carried on glass, or on thin sheets of metal,
to become reticulated under chemical and heat treatment.
The gelatine receives the photographic image direct.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=64 p:43 ]]

It reproduces the whole range of tone values of the
original accurately. Dr. G.B. Harrison reports that
"at its best collotype can be so perfect as to need
a magnifying glass to detect the difference between
original and counterfeit; ink can be exactly matched,
and every spot, smudge and shade of color exactly
reproduced. One of the most brilliant examples of
collotype used for book reproduction was the work of
the Oxford Press some thirty years ago, in their
facsimile of the first Folio of Shakespeare's Plays,
and I would call attention to the show-through at the
back of the portrait as an example of what can be
done with this process. Where funds allow, collotype is
for every purpose the best." (G.B. Harrison: "Facsimile
Reprints", a paper read before the Anglo American
Conference of Historians, 13-18 July, 1931) But the
cost of collotyping is at least three times that of
photo-lithography; it should be classed with art work
rather than printing.

In the reproduction of manuscripts the
photograph will bring out shades which the photostat
would suppress. In copying illuminated manuscripts
the advantage of the photograph is especially
important. Cost and qualitv in both these processes

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=65 p:44 ]]

are very high.

III. Comparison of the Different Methods of Book 
Manufacture

The four fundamental techniques described
above: - raised type printing, photo-lithography,
mimeographing and photostating, - are all competitors
with each other for the production of books. There is
another technique - filmslide photography - which does
not produce books at all, but which offers to the
scholar a tiny roll of film as a substitute for a book.
The filmslide is the competitor, not of any particular
book production technique, but of the book itself as an
instrument for the preservation of records. Before
evaluating the filmslide process it is necessary,
therefore, to review the process of book production as
a whole.

The technological and financial variables in
book production resolve themselves into five, which
should be defined relative to any item which it is
proposed to reproduce:

1. Is the material ready for photographic
reproduction, or does it still require to
be copied on the typewriter? (Newspapers,
printed books, manuscripts of which
facsimiles are desired,and clean typescripts
can be considered as ready for the camera.)

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=66 p:45 ]]

2. Will the purchasers and readers demand book
type and justified right hand margins, or will
they accept typewriter face and uneven right
hand margins? (Bibliophils and scholars will
probably disagree on this question.)

3. Does the material include other things than
straight textual matter which can be tapped
off on a typewriter or linotype machine?

4. What degree of legibility is required?

5. How many copies can be sold.

These variables are all inter-dependent.

1. Reprinting Material which is Ready for Photography.

When material is ready for photography, and the
problem of the use of typewriter face type is not in-
volved, the photo-offset process has no rival. The small
press is preferable for editions less than one thousand,
the large press (such as the Harris) is able to compete in
larger editions, if labor and business policies are the
same.

There are two critical sizes of editions in the
use of the photo-lithographic method of copying books and
newspapers. There is the low limit to an edition, below
which the photostat is preferable; there is an optimum
size of edition which will take maximum advantage of the
falling unit costs of production while avoiding sales
risks and rising costs of promotion.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=67 p:46 ]]

The dividing line between photostat and photo-
lithography in a business establishment which uses both
methods is somewhere between fifteen and twenty copies.
The figures quoted for commercial work are in agreement
with this conclusion.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=68 p:47 ]]

COMPARATIVE COSTS OF PHOTOSTAT AND PHOTO-LITHOGRAPHY IN EDITIONS OF
LESS THAN ONE HUNDRED. 100 PAGES, SIZE 8-1/2 x 11 INCHES. UNIT COST.

[[ Table ]]

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=69 p:48 ]]

The dimensions quoted by the different photo-offset
presses differ because for each size of metal sheet
there are certain sizes of imprint which can be made
most economically, since they divide the metal sheet
without wastage. The Rotaprint 17 X 11 inch sheet
makes two pages 8½ X 11 inches; the big Planograph sheet
makes 16 pages 9 X 11½ inches, or 9 pages 12 X 15½
inches. The cheapest photostat is passed by the most
expensive photo-lithograph at 50 copies; the most expen-
sive photostat is passed by the cheapest photo-lithograph
at 17 copies. This latter figure agrees with the office
practice cited above, of changing from one process to the
other between fifteen and twenty copies.

These figures have practical application to a
present problem in connection with certain projects for
copying old newspapers by the photostat method. The
most successful enterprise of this kind was the one
promoted by Mr. Worthington Ford, who organized the
copying of many colonial newspaper files. In most of
his undertakings the number of subscribers ranged from
fifteen to twenty two, and the list never got beyond
twenty four. This list therefore stood at the marginal
point where photo-lithography was Just beginning to be

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=70 p:49 ]]

superior to the photostat. Projects similar to that of
Mr. Ford are under way on behalf of the John carter
Brown Library at Providence, the New York Historical
Society and the University of Michigan. A new project
is being launched by the Missouri Historical Society,
and an attempt is being made to reproduce the unique
Chillicothe Gazette at the Ohio Historical and Philoso-
phical Society. When subscription lists to these
projects rise above twenty, the photostat should be
replaced with the offset press. five or ten extra
copies could then be made for later sale without in-
creasing the amount the subscribers would have to pay
for their sets. Twenty is the critical point not only
for transition from one technique to another, but also
for the lowering of unit costs by increasing the
edition. A subscription list of forty can purchase
their copies at a unit cost only half as great as must
be paid by a subscription list of twenty. Promoters
might make use of this fact by fixing their maximum
price at the photostat level but offering the advantage
of lower prices in the event that the subscription
list grows large enough to warrant reductions.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=71 p:50 ]]

The other critical point in the photographic
copying of books is the size or edition at which the
unit cost of the book equals the normal prices which
libraries are in the habit of paying for books. This
is between 50¢ and $1.80 per hundred pages. Two enter-
prises which illustrate the importance of this point
are the project of the Facsimile Text Society and the
dictionary project financed with Carnegie funds under
the direction of Professor Fries of the University of
Michigan.

The Facsimile Text Society has 400 charter
members; it has printed photo-lithographic copies of
5 books. It has 11 more books in preparation, and 104
titles under consideration. The cost level at which
it operates is explained in a letter from the secretary:-

"The cost averages about five dollars a page
for one thousand copies, printed on rag paper and
bound in boards. In order to pay their way these
books must sell at least three hundred copies."

This is a unit cost of 50¢ per 100 pages. The catalogue
shows that the price to non-members is about $1.00 per
hundred pages, to members about 63¢. The apparent dis-
crepancy in these figures is probably explained by the
element of bindery cost, which would not necessarily be

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=72 ]]

incurred for that part of the edition which is printed
but unsold. As has been noted in another place, the
work of this society is done on one of the large presses
which cannot store its metal plates for a second run.
The British Shakespeare Association Facsimiles are
printed by a similar method at a similar cost - editions
of one thousand at a cost of 30¢ per 100 pages, stitched
but not bound. These societies print more copies than
they need to sell.

A different policy is followed in the dictionary
project at the University of Michigan, which calls for the
copying of 400 titles in editions of 200. The unit price
is to be 50¢ per 100 pages, and the metal sheets are to be
stored, so that later runs can be made at a unit cost of
about 50¢ for a second printing of 100 and 22¢ for a sec-
ond printing of 500. The binding will probably be less
expensive than that which is used by the Facsimile Text
Society.

The significant fact is that, in editions of
from two to three hundred, reprints of books can be
supplied at costs substantially below the prices which
libraries usually have to pay for average books in the
open market. Even copies of old newspaper files could

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=73 p:52 ]]

be sold at these low prices if the subscription list
could be brought to the two hundred mark. No book
desired by two hundred purchasers need ever be out of
print.

3. Typewritten Books.

When the possibility is once presented that
books may be printed directly from a typescript, the
prospect is opened for the re-distribution of labor
between author and publisher. In this re-distribution
there may be an opportunity for great economies.

In preparing copy for a printed book, the
author is expected to provide a reasonably clean copy.
But the editorial staff of the publishing house always
edits the typescript before it goes to the linotype
machine. The proof must then be corrected to the copy.
This editorial service accounts for twenty per cent of
the manufacturing cost of the book. The author must
then read his proofs, usually twice, once in the galley
and once in the page.

How are the author's duties changed if the
book is to be taken photographically from his type-
script? He has the responsibility of supervising the
preparation of a copy that is final and correct in
every detail. No editor or proof reader will help

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=74 p:53]]

him with his spelling or systematize his capitalization.
He can easily make changes and corrections in his master
copy by pasting new matter on top or the old. But once
he sends the sheets to the press, no further changes can
be made. This has some advantages, for the labor of
reading galley and page proofs is eliminated. The type-
writing in such work must be done with very clean type
and a special dark ribbon. Erasures and changes can be
made if necessary. A part of a page can be pasted over
the original text if an extensive change must be made.
Footnotes can be cut out and pasted at the button of
each second page. A clean typewriter, a dark ribbon,
rubber cement and shears are all the tools needed in
preparing typescript for reproduction. The author has
a free hand to include as much illustrative material or
as many fac-similes as he wishes, or he may insert
Chinese or Greek characters, - for the cost or publish-
ing is not affected thereby. The book which is published
by photographing a typescript follows the same cost curve
as the photographic copy of a printed book.

The devices related to the mimeograph distribute
the labor of printing more as is done in the publishing
trade. The author was the extra labor and the opportunity

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=75 p:54 ]]

for correcting his script which proofreading affords. He
cannot without great extra trouble reproduce anything
which is not on the typewriter keyboard, just as if he
were dealing with a linotype process. These methods
resemble photo-lithographing from typescript chiefly in
that they still leave the right hand margin ragged.

When the matter to be published contains tabula-
tions, characters not in the Roman alphabet, diagrams or
mathematical formulae, the advantage of photographing the
typescript increases because the cost of typesetting is
much higher. Mr. Charles G. Proffitt of the Columbia
University Press writes:

It has long been our judgment that much scholarly
work, now printed by the conventional method of
typesetting, could be produced in a way that
would be just as useful to the small circle of
readers involved and at a price substantially
lower than now paid, if the work were done by the
offset process from clean typewritten manuscript.
We published such a volume some time ago (Westermann
and Kraemer - "Greek Papyri") and are urging this
same method in the publication of one or two other
volumes at the present time. It is particularly
appropriate where the use of foreign characters
makes the type-setting unusually expensive.

Anything in the manuscript, pictorial or otherwise, which
cannot be pounded out on the linotype gives rise to costly
hand work, and possibly to charges for zinc cuts or copper
plates.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=76 p:55 ]]

The redistribution of labor between author, editor.
and printer together with the elimination of special hand work
in the press room account for the greatest cost differences
between the two methods of printing. There is another element
which may mean much or little - the fact that photo-lithography
operates regardless of the number or words on a page, whereas
linotype cost increases directly with the number of words.

Two distinct typographic problems here converge - the artistic
problem of the value of the straight right hand margin which
the linotype will supply, and the legibility problem which
arises when advantage is taken of the ability of photo-
lithography to increase the number of words on a page by
photographing typescript on a reduced scale.

At what price does the publisher purchase the
aesthetic satisfaction of the justified line of type? What
is the cost differential between the two process? Print-
ing costs are higher in America than abroad, but so are
photo-lithographic costs. Foreign printing is a little
cheaper than the cheapest American photo-lithography. J.
Gamber, Paris, prints 50,000 words, 100 pages in an edition
of 200 at a unit cost of 44¢; Edwards Brothers will make a
book of the same size by photo-offset for 52½¢. If 600
words instead of 500 are put on an Edwards page, the cost

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=77 p:56 ]]

per word of the two methods becomes the same. It is
more reasonable to compare the Edwards figure with the
lowest printing quotation in this country, which, for
small editions, is offered by the Mime-O-Form plant.
In editions of 200 to 300 its costs are exactly double
those of photo-offset from author's master copy. To
change from a photo-lithographed typescript edition of
two hundred to the cheapest book printing of equivalent
legibility available in this country requires that
either the unit price or the size of the edition be
doubled.

The Cost of Legibility.

In any photographic process which permits the
enlargement or reduction of the size of copy - that is
to say, in photostating end photo-lithography - the
legibility of the product is one of the variables which
is strictly under control. The printer in these cases
is concerned with a size of page, not a number of words.
The cost per word is therefore dependent upon the
editor's choice of a size of print, and diminishes by
geometrical progression as the point size of the letters
is reduced. The variation in costs is illustrated if
the full size typescript page be compared with n news-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=78 p:57 ]]

paper reduced to half size. The typewritten thesis, now
the standard form in which unpublished scholarly work is
presented, contains about 300 words on a page 8½ x 11 in-
ches. This is exactly the size of a page which will take
a half size reduction of a standard newspaper. But the
newspaper contains 5,000 to 6,000 words. Therefore the
cost per word of photographic reproduction in one typo-
graphic form may be only one twentieth of the cost in
another form.

The problem of the legibility of small type
has not been thoroughly explored in its relation to the
photographic methods of reproduction. Much more atten-
tion has been paid to display type for advertising. A
study of the effect of photo-lithographic reduction on
the legibility of newsprint was made by Professors Don-
ald G. Paterson and Miles A. Tinker of the Psychology De-
partment cf the University of Minnesota, at the request
of the Joint Committee on Materials for Research. Their
curve of diminishing legibility drops rapidly after the
size of the type passes below a 50% reduction, equivalent
to a 3½ point type face. There is no statistically sig-
nificant retardation of speed of reading when the news-
print is reduced to 80% of its full size; at 50% reduc-
tion the retardation amounts to approximately half a word
a second. Fatigue in reading is probably proportional to

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=79 p:56 ]]

retardation of speed. The following table presents the
conclusions covering legibility, number of words per
square inch, and cost of publication.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=80 p:59 ]]

THE COST OF LEGIBILITY

(Page 8½ x 11 inches calculated for printing surface of 6-7/8
x 9-3/8 or 66 sq. in.; Page 5½ x 8½ or 5-7/8 x 8½ calculated
for printing surface 4½ x 7½ or 34 sq. in. Newsprint figures
found by actual count; linotype average figures taken from Hoch:
"The standard book of estimating for printers" issued by United
Typothetae of America. Typescript estimates by Edwards Brothers)

LINOTYPE STYLE.
Size of type

Words
per sq.
inch

Words
per
5½ x
8½
page

Words
per 8½ x
11
page

Unit Cost of
50,000 words
at 1¢ per
page *
SMALL
PAGE
LARGE
PAGE

Legibility
compared
with 7 pt.
newsprint
in words
per second

7 pt. newsprint
reduced 50%
(3½pt) 130 4424 8581 $.12 $.07 .45
4½ pt. set solid 82
5 pt. " " 60
4½ pt. leaded 57
5½ pt. solid 52
6 pt. solid 47
5 pt. leaded 44

7 pt. newsprint
reduced to 80%
(5½ pt) 44 1496 2904 $.33 $.17 .05

7 pt solid 38
6 pt. leaded 34
7 pt. newsprint 33 1122 2178 $.45 $.23 (Standard: 4.3 to 4.5 words per sec.)
8 pt. solid 32
9 pt. solid 27
8 pt. leaded 23
10 t. solid 21
9 pt. leaded 20 680 1320 $.74 $.38 Not tested
11 pt. solid 17
10 pt. leaded 16
11 pt. leaded 14
12 pt. solid 14
12 pt. leaded 11
Pica Typescript
single space 10 340 660 $1.48 $.76 Not tested
double space 5 170 330 $3.53 $1.52 Not tested

* (Small page can be printed for 1¢ in editions of 100; large page in
editions of 200)

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=81 p:60 ]]

These figures take no account of a number of important
factors, such as length of line, and differences in type
face which affect legibility. An experiment intended to
supply legibility figures for typescript is under way at
the University of Michigan. The table here given indi-
cates the existence of a number of point sizes between 7
point type and 3.5 point type where reduction in cost
moves faster than reduction in legibility as expressed
in speed of reading.

Cost and Legibility in Percentages of Newsprint
Cost Reading Speed

7 pt. newsprint.........100%..........,....100%

5½ pt. (7 pt.
reduced to 80%
of size)................78%................98%

3½ pt. (7 pt.
reduced to 50%
of size)................23%................89%

7 pt. reduced to
30% of size -
too small for
practical use............9%................25%

The standard typewriter equipment is such that
directions for preparing typescript for photo-lithography
cannot assume the availability of any other type faces
than the 12 point Pica and the 10 point Elite. But by
using these type faces and reducing them photographically,
it is possible to produce finished copy in sizes all the
way from 4½ points up, in a great variety of styles as to
length of line and spacing. Copy prepared on ordinary
typewriter paper, 8½ x 11 inches, will produce a final

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=82 p:61 ]]

page having anything from 150 to 1500 words, or more.
The normal ten inch typewriter can also be used to
prepare copy upon special large sizes or paper, thus
permitting even greater variety in final size and spac-
ing of print.

The small photo-offset press, which is best
for small editions, works most economically upon either
of two sizes of book page - 8½ x 11 or 5½ x 8½, (some-
times 5-7/8 x 8½). Typescript should therefore be
prepared for reduction to one or these two sizes.

There are three ways of preparing copy for the
8½ x 11 page.

1st - to make four pages of ordinary typescript
and have them reduced to half size as two
columns of a page;

2nd - the typescript can be copied without reduction
in size;

3rd - the typescript can be made in one or two col-
umns upon a larger typing surface and then
reduced to the 8½ x 11 size.

Typescript for the 5½ x 8½ page is easily prepared on
standard typewriter paper for any amount of reduction.

The ratio of length of line to size of letter
has an important effect upon legibility. A good length

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=83 p:62 ]]

is N x 30 where N is the point size of the type. it this
rate the optimum length for a line of Pica type is about
five inches, and of Elite type four inches. If the ordi-
nary double spacing is wider than necessary, and single
spacing too crowded, another spacing is given to a type-
script by using a special ratchet which can be attached to
the typewriter to set the lines one and one third spaces
apart. Each combination of length of line, size of type,
spacing and reduction ratio yields its result in number of
words per page, and hence in cost of reproduction. The
costs are here tabulated; the legibility factor has not yet
been worked out. The styles which are marked with a star
are reproduced in the appendix.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=84 p.63 ]]

PREPARATION OF TYPESCRIPT FOR PHOTO-LITHOGRAPHY

*=sample in appendix

A. Using standard size type- 
script paper 8½ x 11 in.,
in standard typewriter

Words per 
page, 8½ x
11 in. in
photo-lith-
ographed
book

Unit cost of
printing 50,000
words at 1¢ per
page (Lowest
quotation requires
edition of 200)
 
Four pages
6~7/8 x
9-3/4 (im- 
posed as 
double col- 
umn on one
page of bk.)

Pica Double 1000 $.50 *
Pica Special 1720 .30
Elite Double 1200 .42
Elite Special 2000 .25 *

Four pages
5 x 7½ in. 

Pica Double 700 .71
Pica Special 1150 .44
Elite Double 800 .63
Elite Special 1225 .40

B. Using special sizes of paper for typescript, either single
or double column style, with typing surfaces as below:

6-7/8 x 9-
3/8 in. 

Pica Special 430 1.16

6-7/8 x 9- 
3/8 in. 

Elite Special 500
Pica Single 500* 1.00 *

6-7/8 x 9-3/8

Elite Single 600

7½ x 10¼ in.

Pica Single 600
Elite Special 600

8-1/8 x 11 in.

Pica Special 600 .84

8-1/8 x 11 in.

Pica Single 700
Elite Special 700

8-3/4 x 11-3/4 

Pica Special 700* .71 *

8-3/4 x 11-3/4 

Pica Single 800 .63 *

9¼ x 12½ in. 

Pica Single 925 .54

8-3/4 x 11-3/4 

Elite Single 1000 (double column) .50

9-3/4 x 13-1/4 Pica Single 1030 .48 *

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=85 p:64 ]]

There ere two inquiries remaining to be made in
connection with legibility. The psychological study should
be carried further to compare typescript and print faces of
various sizes and densities till all the unnecessary varia-
bles are excluded, three or four levels of legibility are
identified, and the maximum number or words per page togeth-
er with the optimum typescript format are determined for each
level of legibility. The users of research material must
then decide what legibility they want in each class of mater-
ial, taking into account such things as probability of fre-
quent consultation, or the likelihood that the reader will
wish to skim the entire text as against the likelihood that
he will merely search for a particular item. The small
type in the dictionary is not fatiguing to the reader be-
cause so little is read at a stretch. In copying printed
material, whether it be in the form of book, periodical or
newspaper, the requirements of the small press will always
favor certain page sizes, and the requirements of the schol-
ar certain type sizes. If a typescript, a duodecimo volume
in 5 point type and a great folio in 12 point type have the
same kind of utility in research, the reprinting should en-
large the one and reduce the other to whatever size has the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=86 p:65 ]]

appropriate legibility. A certain rule of thumb accuracy
in making decisions about size of type can be achieved
without setting a standard, but standards of legibility
if defined would aid both editors and promoters. The
editor would know at once, when he had decided upon a
standard of legibility, what kind of printing technique
would fulfill that standard. And the promoter would be
able to inform prospective subscribers accurately upon
this important aspect of his project. In connection with
the copying of newspaper files the question is already
presented. Some librarians want full size copies, and
others believe that the pages should be reduced. But
how far is the reduction to be carried? There is no ad-
vantage to choosing a fixed percentage and reducing
everything by that amount, for some newspapers are close
to the low limit of legibility in their original size.
The promoter should be able to say that the copies to be
furnished will have a certain legibility, so that the
subscriber will know exactly what he is subscribing for.
The prospective users should have an opportunity to say
whether they would prefer to have one newspaper file
made accessible to them in seven point type or four files
in three and one half point type.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=87 p:66 ]]

Optical Aids to Legibility.

When the size of the type face is cut down in
order to reduce the price of a volume or ut down the
number of copies which must be sold, the point is ulti-
mately reached where reading glasses may be useful to
assist the naked eye. The optical principles of the
reading glass are such that every increase in magnifying
power diminishes the area of the field of vision. No
practicable magnifying system will do more than double
the ostensible size of the type without introducing disa-
greeable distortion at the edges of the field, or unduly
diminishing the field of vision. There is no such thing
as a really "high power" reading glass. The most conveni-
ent kind of optical aid to the reading of small type is
the half-cylinder of glass which is mounted on short run-
ners which hold it about an inch above the page. This
apparatus magnified the letters only in the dimension of
height, but lends itself more easily than the round read-
ing glass to the rapid skimming of a page. When the
limitations of the reading glass are considered in connec-
tion with the technical difficulty of reducing print to
sizes much below 3.5 points, it is clear that there is no
advantage in making a text which cannot be read with the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=88 p:67 ]]

naked eye. The aid of the reading glass and its variants
should be saved for the use of those whose eyes are sub-
normal, or who find themselves obliged to read continuously
in print which is only suited for short stretches of read-
ing. A reader who merely looks up an article in a 3.5
point newsprint page will not need optical aid, but a man
who is going through an entire file of the newspaper may
have use for one of these glasses.

The fiskoscope

The fiskoscope, invented by Admiral fiske, is a
magnifying glass intended to render possible the reduction
of print to the point where reading with the naked eye is
impossible. Typescript or other reading matter is reduced
seven diameters in the offset press and printed in thin
strips. 3 x 8½ inches, each strip containing 45 small
squares of 165 typescript words each, a total of over 7400
words, equivalent to maximum number of 50% reduced news-
print words which the offset press can put on a page 8½ x
11 inches. The strip of finely printed paper is fixed in
a movable frame in front of the magnifying glass, the
glass is set at a fixed focus which suits the reader's eye,
and the printed matter is then moved through the frame by

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=89 p:68 ]]

turning a little screw.

To a person not trained to use one eye alone.
the fiskoscope is very tiring. It is difficult to use in
reading normal pages because the field of vision is too
small. The special pages of 165 words each were type-
written to permit the whole page to be seen at once, but
in reading a page of the New York Times, reduced 6 diame-
ters, the restriction of the field of vision precludes
skimming and is very annoying. If the principle of the
fiskoscope is to enter usefully into the scholar's equip-
ment, it must be keyed in with filmslide photography
rather than with offset printing, for it calls for a
departure from the book form. Admiral fiske writes that
the limit upon the magnifying power of his instrument is
fixed by the printers, who cannot provide him with copy
any smaller than newsprint reduced six diameters. But
the filmslide negative will reduce newsprint more than
twenty diameters. The place for such an instrument as
the fiskoscope is as an auxiliary to the projector in the
use of filmslides,-- a use which will be discussed below.

Quality of Paper.

In small editions the cost of materials consti-
tutes so small a proportion of the total expense that the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=90 p:69 ]]

quality of the paper can be improved without greatly
changing the price of the books.

The problem of permanence in book papers is
a complex chemical problem now being studied by the U.S.
Bureau of Standards and by a Special Committee on the
Durability of Paper set up by the British Library Associa-
tion. The tentative conclusions arrived at in the Bureau
of Standards have been set forth in a paper by B.W. Scribner
before the American society of Mechanical Engineers. The
British committee has printed its report in a pamphlet
available at its office, 26-27 Bedford Square, London,(ls).
In both investigations it was recognized that several de-
grees of permanence in paper must be standardized. The
British classification is as follows:

Grade I. For use where "absolute" permanence is
required.

Grade II. For use where relative permanence is
desired and where a competitive price is
necessary.

The U.S. Bureau of Standards classification runs:

Grade I. Permanent papers, having a maximum degree of
purity.

Grade II. Papers quite highly purified, that may be
expected to have a minimum life of 100 years.

Grade III. Papers quite highly purified, that may be
expected to have a minimum life of 50 years.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=91 p:70 ]]

Grade IV. Papers having a low degree of purity, for
current use only.

Grades I and II are distinguished in the English
report by the nature of the raw material used and the price
level at which they can be furnished, in the American Bu-
reau of Standards report by a computed life-probability and
performance in oven-aging. The American Rag Paper Manufac-
turers hold with the British Library Association in favor
of raw materials as the criterion of probable durability;
the U.S. Government Printing Office uses Bureau of Standards
"performance" specifications, which are "revolutionary" in
business practise. In the following tables of specifica-
tions it is to be noted that "Alpha Cellulose" is derived
either from rags or wood pulp. The opposition between the
two methods of grading, by raw-materiel composition and by
performance, is clear:

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=92 p:71 ]]

CHEMICAL AND PERFORMANCE SPECIfiCATIONS

(n.s. = not specified)

I II

U.S. 
Bur- 
eau 
of 
Stan- 
dards 

U.S. 
Rag 
Paper 
Mfg. 
Assn.

Brit- 
ish
Lib. 
Assn. 

U.S. 
Bur- 
eau 
of 
Stan-
dards

U.S. 
Rag 
Paper 
Mfg. 
Assn.

Brit-
ish
Lib
Assn.



Alpha
Cell-
ulose 

90% n.s. n.s. 80% n.s. n.s.

Raw 
Mater- 
ial 

n.s. 100% of new rags 100% of new rags n.s. chem. wood pulp chem. wood pulp

Acid-
ity 
Above 

5 pH 5 pH 5½-6 pH 5 pH 5 pH n.s.

Copper
Number
Below 

1.5 0.5 n.s. 2.5 1.5 n.s.


Rosin
at
least 

1% 1.2% n.s. 1.5% 1% 1% or 2%

Mineral
Loading 

n.s. None None n.s. n.s. 5%

Perform
mance 

see below n.s. n.s. n.s. n.s. n.s.

Performance test for Grade I paper (U.S.Bureau of standards)

Performance test for Grade I, :  changes on heat-
ing 72 hr. at 100 deg. cent.; decrease in content of alpha
cellulose should be not more than 1.5%; decrease in fold-
ing endurance not more than 25%, increase in copper number
not more than 0.5%

Mr. E.H. Naylor of the Rag Content Paper Manufactur-
ers Association defends the "composition" specification, and

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=93 p:72 ]]
72

questions the validity of the tests of permanence proposed
by the Bureau of Standards. He writes that "there is not
today any test or set of tests which adequately describe
the permanence of paper", and offers further concrete
criticisms:

"Alpha cellulose content. This test measures the
amount of papermaking fibrous material as distinct from
cementing material and definitely degraded material and
impurities. It takes no account of the wide difference
in quality of fibrous materials such as is indicated by
the test for viscosity of cuprammonium solution. Papers
from the best of wood fibers have an alpha content of
around 90 per cent, while those from new rags have as
high as 98 per cent. Therefore papers in Grade I may be
made of the best of new rags, degraded or worn rags, the
best of purified wood pulps or a mixture of good rags
and an inferior wood pulp. Price and permanence quali-
ties within this range will necessarily vary widely.

"Copper number. The best of wood pulp normally
carries a copper number several times as large as that
of similarly purified rag stock. While new rag papers
show copper numbers of .15 to .75, wood pulp papers vary
from 1.0 for alpha pulp to 2.5 for ordinary sulphite.
The range of permanence possibilities for papers of
copper number less than 1.5 is even wider than that of
the alpha content specified....

"Oven aging accelerates the action of certain im-
purities in paper and indicates in a conveniently short
time the possible slow degradation of paper from such
impurities. At the temperature chosen this dark heat
has no effect on pure fibers whether of wood or rag and
it gives no indication of their relative resistance to
atmospheric impurities. At a temperature of 135 C
instead of 100 C rag papers are decidedly more perma-
nent than wood papers."

Mr. Naylor declares that tub sizing with gela-
tine adds a protective feature by increasing resistance

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=94 p:73 ]]

to light and atmospheric impurities.

It must be concluded that the scientific class-
ification of papers according to permanence is still a
highly contentious subject, in which the dispute is ag-
gravated by the rivalry of the wood-pulp and rag-content
interests.

Although some of the rag paper manufacturers
will not bid on the Bureau of Standards specifications,
American Writing Paper Company quotes 35¢ a pound for
Grade I, and 22¢ a pound for Grade II. The cost of quali-
ty is figured differently on the British scale. Taking
normal book paper prices as 1, their quotation for the
Grade I papers is 3 to 6, and for Grade II papers 1.

The American Rag Paper Manufacturers suggest a different
price index in a letter of November 17, 1931, which is of
special interest because it expresses the considered opin-
ion of the industry:

A meeting was held with representatives of
the various mills making permanent book
papers in order to discuss the request in
your letter of August 16th for the differen-
tial in printing cost between printing on
ordinary book stock and on stock having the
higher degree of permanence. Those present
were unanimous in their expression of inte-
rest and desire to cooperate to the extent
of their ability....

These manufacturers have felt that our

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=95 p:74 ]]

present knowledge of the factors which
enter into paper permanence is not
sufficiently well defined for them to
adopt an adequate uniform set of spec-
ifications, and therefore the permanent
book papers offered by these mills diff-
er as to details of quality and price....

Some papers made of cheap, impermanent
materials may command a reasonably high
price because of some attractive feature
as color, surface, finish, bulk, weight,
printability, etc., while the less attrac-
tive permanent papers are selling for less
price. A price limit which a buyer might
construe as defining the permanence of
paper could obviously lead to serious
error.

A price limit, however, together with a
few carefully chosen specifications which
will qualify the price as being due to
permanence quality rather than some ap-
pearance effect they consider as highly
desirable...

There follows a discussion of the permanence specifica-

tions, leading to the conclusions charted above.

By mixing rag with chemical wood, both being
of the quality indicated above, an inter-
mediate grade with prospects of life longer
than for Grade II may be predicted.

If we take 1 as the equivalent price prevail-
ing today for the ordinary lower quality of
printing paper and for papers of Grade II
above, the ratio of price for the various
grades above would be about as follows:

Grade I.........4
Grade II........ 1
"Mixed furnishes", 2 to 3.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=96 p:75 ]]

It goes without saying that research materials
of permanent value should be printed upon permanent pa-
pers. How much does the cost of permanent paper increase
the cost of book production? Two factors enter into this
calculation,- first, the extent to which an edition is
sold out, and second, the size of the edition.

If the weight of paper in a book is one pound,
an edition of 1000 will use paper as follows:

Cost of
Paper

Class I, British or U.S.
Bureau of Standards $350 (American Writing
Paper)

Class II, U.S. Bureau of
Standards $220 (American Writing
Paper)

Class II, British $85 (American Rag Paper
Mfg. Assn. Ratio)

A paper cost differential or 7¢ per volume if
the whole edition is sold out will rise to 43¢ per volume
if the edition is expected to pay its way with the sale of
300 copies. A paper cost differential of 26½¢ per volume
if the whole edition is sold out will rise to 88¢ if only
300 copies of the edition are sold. It is the storage of
expensive paper in unsold books that makes it so costly to
print on high priced paper. The financing of permanent
paper reduces itself to the problem or how to avoid print-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=97 p:76 ]]

ing more than can be sold. In offset printing this means
that the key to paper quality is the use in printing of
a metal sheet that can be stored for later printings.
This applies to all non-commercial enterprises.

In ordinary commercial publishing the risk of
printing a large number of copies that will not be sold
must be taken, for the possibility of profit depends upon
the smaller unit costs of books in large editions. The
practicable way to finance high quality paper in books
published commercially would appear to be in the print-
ing of one or two hundred copies of fine paper for sale to
libraries. But such printings, when made, are usually
sold at luxury prices. The publishers say they cannot
afford to stock two qualities of paper at approximately
the same price because of the extra stock-room cost. It
is possible that libraries could secure a block of books
printed on special paper at no higher cost than the added
cost of the paper if they were organized to give the pub-
lisher the order in one block, so that there would be
no stock-room cost. Any organization of libraries which
had purchasing power sufficient to reprint new editions
at will could probably arrange to secure special paper in
many commercially printed books.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=98 p:77 ]]

The influence of size of edition upon the cost
of using higher priced papers is very marked, for the
smaller the edition, the smaller is the proportion of
its cost due to paper. This fact reinforces the argument
that small editions printed with the certainty that all
copies will be sold can be produced upon the most dura-
ble paper without greatly adding to their cost.

Recommendation on Book Production.

The interests of research in the reproduction
of material are of two kinds, an interest in the present
and interest in the future. From the standpoint of the
present it is important to give the scholars of today an
equipment which they can use with greatest economy of
effort. The disadvantage of having material reproduced
in small type must be balanced against the advantage of
having it available on the spot. The sacrifice of artis-
tic values in the use of typescript must be weighed
against the gain of a greater freedom from financial
restrictions in publishing research work and material.
The further research proceeds in the direction of spec-
ialization, the more limited will be the circle of
readers interested in any particular item of research

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=99 p:78 ]]

material. Unsubsidized publication has hitherto financed
itself on the possibility that the circle of purchasers
would reach two thousand; technological conditions permit
a new basis - the certainty that the circle will reach
one hundred. Scholars in any field of study in the social
sciences and humanities which can organize their group of
one or three hundred purchasers who are able to pay normal
book prices for materials in the field can escape from
financial limitations upon the circulation of any materials
that are deemed important.

The same organization of purchasing and produc-
tion would also care for the interests of the future, for
publication on permanent papers could be guaranteed.

It is not to be expected that there could be
found one hundred institutions, each able to subscribe for
all worthy ventures in publication by and for scholars.
Perhaps not more than five or six institutions in the
country have the resources which would permit them to
specialize in everything. For libraries themselves must
specialize and farm out their work, and leave it to lib-
eral lending policies to make up the difference between a
particular equipment and a particular need. The group of
subscribers who will fill in files of reproductions of

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=100 p:79 ]]

Colonial newspapers will probably not interest themselves
in reproductions of contemporary European newspapers;
those who went to buy a copy of every reprinted book of
16th century English may be indifferent to the problem of
filling in the gaps in the collection of the statutes of
the American States. As things stand at present, the
organization of the market has lagged far behind the
development of technical processes of reproduction.

The great libraries, - such as Harvard, Library
of Congress, New York Public Library, - will not always
be the purchasers of reprints of old materials, for they
may be in possession of original copies. Younger and
poorer libraries may often be the ones which provide the
purchasing power for a reprinting enterprise. As an
illustration of this situation, the Miller Diary can
serve. This is a set of twenty volumes of sources on the
Paris Peace conference. It was printed in an edition of
40 copies only - a restriction which is regretted by Mr.
Miller, who financed the publication and gave the volumes
to libraries free of charge. Only a few of the libraries
in America possess this important set. It is technically
possible to reprint it in an edition of 300 to be sold to
other libraries at $2.50 per volume, or fifty dollars the

[[ this page was missing p:80 ]]

set. The purchasers would in this case be the medium
sized libraries which were not able to secure a copy of 
the original edition.

Organization of purchasing groups in their 
respective fields of interest is the next step in liberat-
ing book production from financial restrictions. But 
there is another possibility - that the book as the ve-
hicle of reproduction for a very narrow circle may find 
itself superseded by the filmslide. The range of useful-
ness of the filmslide will be considered in the next
chapter.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=102 p:81 ]]

IV. filmslide Photography

The reduction of a text to microscopic propor-
tions on a photographic film or plate may enter into two
entirely different schemes for the reproduction of research
material. It may be a means of imposing a negative in the
making of a full-size photostatic copy, and hence a mere
variant of book-manufacture by the photostatic process.
Or it may be a device which is independent of book manufac-
turing, and which will put before the scholars' eyes an
image of a text projected on a screen, rather than the
tangible page of a book.

When the filmslide is used as a step in the pho-
tostat process the technique of photostating is changed,
but the cost level is still about the same. The cost of
the negative is minimal, the apparatus for taking the neg-
ative is light and easily portable, and it is possible
to store the negative for a long period and then enlarge
it as a full size positive print when there is need for it,
or when money is available. This is the way in which Mr.
Worthington Ford plans to use filmslides in collecting
copies of foreign archive material for the Library of
Congress. But in this project the making of filmslides
has far outrun the work of transferring them to full size
sheets of photostat paper. In the meantime the film-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=103 p:82 ]]

slides themselves are being read by the projection meth-
od. It is this latter process that promises the most
sweeping changes in the cost level at which research
material can be reproduced.

All types of equipment now available for film-
slide reproduction of materials make use of standard
positive motion picture film, 35 mm. wide. Although
there are experiments in the motion picture industry
looking toward the use of wider film strips, this size
must for the present he assumed to be standard. There
are two kinds of material out of which the film is
made - a nitrate and an acetate base. The former is
considered more perishable, and is certainly more in-
flammable. The acetate film is the only kind which
should be used. Until recently it was sold by Eastman
for 2¢ a foot, then Dupont cut the price to 1¢, and
Eastman came down to 1½¢. The size of the surface of
the film which can be used depends upon whether it is
perforated or unperforated. If the film is perforated
to meet the mechanical requirements of moving picture
cameras and many other kinds of apparatus, only 25 mm.
of the strip is available for photography. If the film
is unperforated this amount is increased to 30 mm. A

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=104 p:83 ]]

square metre of sensitive surface is furnished by 108
feet of unperforated film, or 132 feet of perforated.
An unperforated strip six feet three inches long, or a
perforated strip eight feet three inches long, will
provide a square foot of photographic surface.

It has been noticed that as the processes
verge from the typographic toward the photographic, the
size of the edition becomes less significant as a con-
trolling element in costs, and the size of the letters
becomes more important. The extreme case of cost reduc-
tion through reducing the size of the letters is film-
slide photography. The square foot of surface, which
costs less than ten cents, will photograph 100 square
feet of print reduced to one tenth its original size,
or 400 square feet reduced to one twentieth. If the
reduction ratio be expressed as a whole number, the
amount of text that can be reproduced on a given quan-
tity of film will increase as the square of the reduc-
tion ratio. A fifty page issue of the New York Times
covers with print two hundred and seventy square feet
of paper surface. It sells for three cents on the
street. Photographed full size it will cover eighty
six dollars worth of photostat paper; reduced twenty

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=105 p:84 ]]

four diameters it will go on three cents worth of film.

The question is therefore raised at the outset:
what reduction ratios are practicable when printed matter
is to be photographed on small film and read by projection?
The physical limit to reduction is certainly the resolving
power of the film. This may be as high as 90 lines to the
millimeter, but is usually 50. (Otto Sandvik, "On the
Measurement of Resolving Power of Photographic Materials",
Proceedings of the Seventh International Congress of Pho-
tography, July 1928) It is affected by a number of varia-
bles, notably the wave-frequency of the light and the
length of exposure, the quality of the emulsion, (especially
the fineness of its grain), and the method of developing.
Dr. Miller Reese Hutchinson has worked out a treatment to
increase the resolving power of film by giving it an extra
bath before developing it. The commercial value of his
invention will probably be exploited in the making of
sound film, where high resolving powers are of supreme
importance. In the development of increased resolving
power, as in its choice of film, filmslide photography is
a very obscure beneficiary of the motion picture industry.

The standard scientific measure of resolving
power is in lines per millimeter. But this proves to be

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=106 p:85 ]]

too refined a measurement for studying the practical
usefulness of apparatus on the market, under normal
shop or library conditions. Since the final term in
the whole process is legibility of material, a rough
index of legibility of product can be used to make
possible the comparison of the efficiency of the diff-
erent filmslide cameras. If a page of seven point
newsprint type is reduced R diameters on the filmslide,
and projected on a screen, the different degrees of
legibility of the screen image can be set apart as
follows:

"A" Legibility - all letters distinct. The pro-
jected image at "A" legibility may be rendered
more easily legible than the original text, for
it can be increased to a larger size.

"8" Legibility - The edges of the letters begin
to blur. The loss in clearness compensates for
any increase in size. But the letters are still
perfectly formed.

"C" Legibility - The tops of the "e"s and "a"s
are blotted over solid, but the openings are still
clear in the "o"s, "d"s, "p"s, and "b"s. Legibil-
ity is distinctly lower than in the original.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=108 p:86 ]]

"D" Legibility. Almost illegible. The openings
of the "o"s, "d"s, etc. are blotted in solid. Sep-
arate words can be deciphered, but not read.

"E" Legibility. Absolute illegibility.

Under practical working conditions the finest
apparatus (Lloyd B. Kennedy equipment) will produce B
legibility in newsprint reduced 24 diameters; the coarsest
apparatus (Ansco) will produce this degree of legibility
at 16½ diameters reduction. The reduction ratio at which
legibility passes from B to C is the limit for practical
work.

The amount of reading matter that can be placed
upon a filmslide depends of course upon the size of the
slide as well as the degree of reduction. The following
are the dimensions of the filmslides in the different
commercial cameras:

Width 
Across 
film

Length Along 
film Strip 

Kind Of
film Used

Ansco 23 mm 18 mm Perforated
Cinescope 23 mm 24 mm Perforated
Q.R.S. " 25 mm 32½mm Perforated
Leica 25 mm 38 mm Perforated
Lemare 30 mm 45 mm Non-Perforated
filmograph 30 mm variable Non-Perforated
Kennedy 25 mm " Perforated
" new style 30 mm " Hon-Perforated

Naturally the size of the page to be reproduced

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=109 p:87 ]]

may bear such a relation to the dimensions of the film-
slide that only a part of the sensitive surface can be
used. A page of the New York Times reduced 24 diameters -
(the limit with the Kennedy apparatus) - can be imposed
on a perforated film strip crosswise, the top of each
page being parallel with the edge of the film. Imposed
in this way the 50 page issue of the Times would require
a little more than three feet of film. With unperforated
film a reduction of twenty diameters would just serve to
lay the page crosswise on the film. There would be no
advantage in increasing the reduction ratio beyond 20 for
unperforated film, or beyond 24 for perforated, unless it
were raised to 36 or 30 diameters, at which point it
would be possible to lay a complete spread of the news -
paper across the film strip. At these higher reductions
the fifty pages would go on as little as sixteen inches
of film, costing less than two cents, but the legibility
under best conditions would drop to D.

Price comparisons: Ansco, Q.R.S., Leica, Lemare, Photo-
scope-Cinescope

Price of equipment for filmslide reproduction
of materials includes three items. The first two are
indispensable: 1) a camera with its copyholder and
stand for copying, 2) a projector for reading. The
third may not be needed: 3) enlarging equipment for

[[ page missing p.88 ]]

transferring a text to a positive print. These diff-
erent elements are combined in different ways. Leica
equipment is made in so many qualities that two alterna-
tive prices must be given for it.

Ansco Q.R.S. Leica (low) Leica (high) Lemare Photoscope-Cinescope

Camera 
with 
Copyboard

$50 $162 $56.36
(3250 fr. plus 30% duty) (1175 fr. plus 20% duty)

Camera 
without 
Copyboard

$57.50 $72

(Leica 
uses two copy
stands, one 
of which is 
accessory to 
the enlarger):

$25 $30

Projector $19.50 $19.90 $49.50 $60.00 $64.80 $34.56

(1300 fr. plus 20% duty) (720 fr. plus 20% duty)

Enlarger $20.00 $55.50 $85.50

(with projector)

The Spencer Lens Company manufactures a projector,
Delineascope B, with special attachment for the
Lemare Camera. It sells for $75, and is better
than the Lemare projector, according to Dr. Martin
of the Library of Congress, who has used both.

The Leica camera will not do short range copying work
without one of the three special front lenses, which cost
$6.00 each.

The minimum equipment for making filmslides and

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=110 p:89 ]]

projecting them costs as follows:

Ansco.................................$69.50
Cinescope-Photoscopie (computing 25%
duty)........... 90.92
Leica (computing 10% academic discount) 146.70
Lemare (computing 25% duty)............226.80

If all useful accessories offered by the Leitz
Company are purchased, the total will exceed $400. The
full Lemare equipment costs $332. The Cinescope figure
does not include a camera or projector stand, but it is
claimed that an ordinary camera stand can be used. None
of these figures include cost of a screen.

The writer has investigated eight kinds of cam-
era equipment for making still pictures out of moving
picture film. Two of these, the Q.R.S. and the Vi-cam,
must be eliminated.

The Vi-Cam takes a long roll of film at one
loading, but has a fixed focus at five feet and is de-
signed for cheap portrait work for registration blanks
or other documents which have photographs attached to
them.

The Q.R.S. is designed for amateur work in
portraits, landscapes etc. It resembles the Ansco Memo
Camera in most respects except that it takes a double
frame picture. It has no special equipment for copying

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=111 p:90 ]]

documents, and will not focus at a shorter distance
than three feet. At this distance it copies an area
15 x 19 inches, reducing it 15-1/2 diameters. With
a better lens and focusing attachment it would be a
very useful camera, for it will take at least fifty
exposures, and sometimes as many as eighty, in one
loading, and it is equipped with an indicator to tell
how many exposures have been used. As the legibility
table indicates below, the Q.R.S. will not make a sat-
isfactory copy of 7 point type at its fixed reduction
ratio.

Cinescope - Photoscope

A Belgian company which began operations sev-
eral years ago sought to sell films "microphote" de la
Photoscopie as a cheap but efficient substitute for
glass slides used in art study and visual education.
It issued a catalogue of geographic, religious, scientif-
ic and artistic subjects, and sought to interest the his-
torical associations and library associations of Europe
in its use as a preserver of pictorial and documentary
material. The Photoscope went out of business, to be
succeeded by La Cinescopie, Rue aux Laines, Brussels.
Then the name became Photoscope again, and the address

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=112 p:91 ]]

changed to 121 Rue Berckmans, Brussels.

There are no Cinescopes in America, a fact
which the company explains as due to the rush of Euro-
pean orders. The description of the camera is given as
"most portable, being only 9½ x 4½ x 6 cm." "The shut-
ter makes possible all exposures". "The actual film
carrier takes a length of three meters standard cinema
film", which means 120 pictures. "A holder containing
more films is contemplated for those who wish to take
as many as 500 or more views on a single film. The
camera can be fastened to any photo-stand". The projec-
tor can be fastened to a stand and used to project down-
ward on a table.

The manager of the Cinescopie wrote in Febru-
ary, 1931 that he would make a test exposure of a page
of the New York Times sent him for the purpose, but he has
not since been heard from. There is no evidence that this
apparatus possesses any superior technical qualities which
should cause it to be ranked above the products of the
Ansco or Leitz companies, which are sponsored by sales
organizations in America, or the Lemare outfit, which has
been amply proved by its work in the collection of archive
copies for the Library of Congress.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=113 p:92 ]]

The Ansco Universal Still-film Copying Camera

The Cinescope was developed to meet a demand
for illustrative material for lectures, and this same
need brought into existence the copying camera sold by
the Agra Ansco corporation, Binghamton, N.Y., for fifty
dollars. One of these cameras was loaned to the Joint
Committee for test and examination. The lens is mount-
ed on a frame facing a copy board which can be clamped
in position at any distance up to three feet from the
lens. The copy board will hold a sheet 11 x 14 inches,
which is the largest size that can be photographed con-
veniently with this apparatus on the single-frame film-
slide (23 x 18 mm.). In back of the lens two chambers
are mounted, one with a ground glass back for focusing,
the other holding the film for exposure. There is no
shutter attachment; exposure is made by removing a
little cap from the lens. In operating the equipment
the original is fastened to the vertical copy board,
the ground glass is moved opposite the lens and the
lens sleeve turned to bring the object into focus. If the
copying is fine, a microscope must be used in adjusting
the focus. When the lens is correctly set, the opening
must be stopped down to increase the fineness of the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=114 p.93 ]]

image, then the cap is put over the lens, and the film
chamber moved behind it. Exposure of 10 to 25 seconds
is necessary with two forty watt electric lights.

The Ansco copying camera is very easy to load.
The film cartridge is simply a wooden box into which the
film has been pushed through a narrow felt-lined slit.
A loading consists of 50 exposures, or three feet of
film. The disadvantage of the loading system as com-
pared with Leica equipment is the absence of any signal
to show when the fifty exposures have been used up.
Another noteworthy point is the cost of the film if it
is bought as recommended in cartridges of fifty expo-
sures. These are sold by Ansco for fifty cents, although
the film in them can be bought for three or five cents.
The user can avoid this charge by loading the little film
boxes himself.

Another difficulty in the use of the Ansco
equipment arises from the vertical position of the copy
board. It would not be difficult to make some kind of a
glass covered book holder with a spring back to hold
book pages flat against the glass, but at present no such
equipment is offered by Ansco. In its absence the diffi-
culty of copving a page from a book is very great. Loose

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=115 p.94 ]]

sheets are clamped in place on the board by means of
some spring steel clamps, but the clamps sometimes
fail to hold. Focusing with a microscope is a very
awkward task. It must be concluded that although the
Ansco equipment is devised for the sole object of
permitting easy and cheap copying of pages, it has
faults which cameras built for more general use do
not have. These faults could be rectified by the
manufacturer without great trouble. The two needs are:

1. A more efficient copy holder. A purchaser
could probably construct this himself.

2. A signalling device to indicate when the
chamber is emptied of film.

Loading the Ansco is easier than loading the Leica.

The advantage of the Ansco as compared with
the Leica and Lemare cameras is to be found principally
in the lower purchase price and the smaller frame of
film used. The latter is an advantage only to the ex-
tent that Ansco focusing will permit the copying of a
page or spread on one frame. Thus if one were copying
a pamphlet 5½ x 7 inches, he could put two pages on
one frame in either Ansco or Leica. But the Leica
frame is twice as large as the Ansco, and the cost of
materials would therefore be twice as high. But if

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=116 p:95 ]]

the pamphlet were larger than 5½ x 7, or if it were nec-
essary to use a high resolving power in making the copy,
the advantage would be with the double-frame cameras 
Leica or Lemare. It is possible, by a somewhat compli-
cated manipulation, to take single-frame pictures even
with the Leica or Lemare.

Leica and Lemare Equipment Compared with Ansco.

Focusing

The Leitz Company has not manufactured a camera
for specialized service in copying. It offers rather a
large variety of attachments which adapt its general-purpose
camera to copying needs. The Ansco also has its general
purpose camera and a way of using it on a copying stand, but
this combination is not as good as the special Ansco Copying
Camera. Therefore the "universal" Leica with its attach-
ments should be compared with the special Ansco camera.

The newest attachment offered by the Leitz Company
(costing $12.00 to $25.00) makes its copying equipment re-
semble the Ansco in that the camera containing film is moved
away from the lens to permit focusing on ground glass, and
then moved back to position behind the lens to permit expo-
sure. But this kind of focusing has little special value
when the image is as small as the image of print reduces to

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=117 p:96 ]]

the size or a filmslide. It is still necessary to use a
magnifying glass in focusing. The standard Leica equip-
ment adjusts the focus by measuring the distance from
camera to object, computing size of object that will be
recorded at that distance, and turning the lens sheath
on a callibrated scale.

Copy Holding.

All the Leica and Lemare copying stands hold
the camera above a copy which lies flat on the table.
The difficulty of clamping a sheet to a vertical copy
stand is avoided. A glass plate can be laid on top of
the object to flatten it. The Lemare copystand is su-
perior to the Leica, especially in its attachment for
photographing pages from bound books. The Leica copy-
stands are of two kinds. There is a light portable one
which is easily dismounted and put into a canvas kit
bag. There is the heavier and more rigid stand used in
common for the copying and enlarging processes. As the
apparatus is offered by the Leitz company, the purchaser
gets his choice between rigidity and portability. Pro-
fessor Gehlke, of Western Reserve University, has con-
trivad an inexpensive modification of the rigid enlarg-
ing stand to make it dismountable and portable.

Double Exposure.

The Leitz Company offers two kinds of cameras:

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=118 p:97 ]]

the one with focal plane shutter at $91.50 and the
"Compur Shutter" at $72. The difference in shutters
is important only in connection with fast exposures,
and does not affect copying work. It happens quite
by accident that the cheaper Compur shutter is better
for copying, because the film winding device works
independently of the shutter, and permits a double
exposure on the same slide. This double exposure
possibility, ordinarily an embarrassment, may offer a
way of saving film by taking two pictures on one slide.
A piece of black felt can be laid down over the area
in focus; a book can be laid down over one half of the
space, and one exposure made, photographing two pages;
then the book can be moved over to the other half of
the space, and another exposure can take the next two
pages. Thus four pages of a pamphlet can be put on one
slide. Used in this way, the double frame camera such
as the Leica or Lemare can do the work of the single
frame camera such as the Ansco.

Loading.

The load taken by the Leica camera is 5¼ feet
of film, 36 exposures, as compared with three feet, 50
exposures in the Ansco; 8 to 10 feet, 50 to 80

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=119 p:98 ]]

exposures in the Q.R.S., and 16½ feet, 100 expos-
ures in the Lemare. In Ansco and Leica the film is loaded
into a cartridge in the dark room, and the cartridge then
inserted in the camera by daylight; the Lemare camera is
taken into the dark room and loaded direct. Convenience is
on the side of the maximum amount of film in one loading,
and hence favors the Lemare.

filmograph and Kennedy Apparatus

It Cinescope, Ansco, Leica and Lemare have been
designed to fill needs more or less closely related to those
of a scholar, teacher or librarian, the filmograph and Ken-
nedy machines have been developed without any reference to
the needs of the academic world. And yet they may turn out
to be better suited than any other equipment to scholarly
uses. In Seattle, Washington, Carl E. Lundberg and H. C.
Lambach have developed a machine called the "filmograph"
for copying court house records for abstracting companies.
Mr. Lloyd B. Kennedy in Warren, Ohio has invented a simi-
lar machine which he uses for the same purpose. Both of
these machines are much larger than the other types, and
are intended to work faster, making longer runs. They
both take full 400 foot rolls of film. The Kennedy appa-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=120 p:99 ]]

ratus is focused by trial and error to pick up the chem-
ical focus, which is different from the optical focus
when very short wave length light is used. The focusing
need be done only once for a job of copying hundreds of
volumes. A volume of records is placed in a glass-covered
copy holder; the attendant presses down the glass cover
which releases the camera shutter. An electric motor opens
the shutter for the exact exposure time required. When
the exposure has been made a bell rings, which notifies the
attendant that he may raise the glass, turn a page, and
press down the cover again. The filmogreph is similar, but
does not photograph through glass, and does not seem to have
so fine a focusing system. Both machines, it is claimed;
work very fast. The filmograph claims that it can make
forty exposures a minute, and copy a 640 page volume of
legal records in twenty minutes. The filmograph which is
set up in Washington, D. C. to copy some land office rec-
ords to replace those destroyed by fire in North Dakota
can make 5,000 negatives, recording 10,000 pages, in a
day. An independent observer has noticed that the actual
daily produce is about half of that - 2,500 exposures,
5,000 pages per day.

The advantage of these larger machines is in

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=121 p:100 ]]

their saving of labor. Dr. Thomas Martin, who super-
vises the use of the Lemare equipment in the Library of
Congress, calculates that two or three operators assist-
ing each other could run off 5,000 prints with a Lemare
camera as easily as one operator could make the 5,000
negatives on the filmograph. The Kennedy machine is
certainly as fast as the filmograph.

The commercial exploitation of these machines
has so far been based upon competition with typescript
or photostat copying for abstract offices. Kennedy
charges 35¢ per foot of finished film; filmograph sug-
gests the rather complicated arrangement that the ma-
chine could be leased for a flat monthly charge of
$50, and in addition to the leasing charge a royalty
charge of 2¢ for each two inches of film, "with the
understanding that we furnish the film at no addition-
al cost to the lessee, but that the developing expense
be borne by the lessee" - a royalty of 4¢ per foot of
film. The going price for developing is approximately
3¢ per running foot. Neither of these arrangements
are suited to the library price level. But it must be
understood that both prices are set in rather arbi-
trary independence of costs, and in ignorance of the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=122 p:101 ]]

price range at which large scale use in research can
be developed.

Both filmograph and Kennedy evince an inte-
rest in developing the research materials field. Mr.
W.K. Shissler, a filmograph representative, writes of

a plan in contemplation whereby the
persons interested will make available
to persons and institutions desiring
such service, what will probably be
known as a filmograph Library Service.
The purpose of this service will be to
make available filmograph duplicates
of invaluable manuscripts in this and
other countries which would not other-
wise be available to persons and in-
stitutions at a distance from the orig-
inal material.

These filmographs with explanatory notes
by noted authorities and custodians would
be classified into various types of film-
ograph libraries to be rented or sold to
those desiring them....Many variations of
the plan might be developed including the
interchange of such filmographs by various
institutions through the filmographing and
distributing service of the organization
which is contemplating this unique service....

It must be concluded that the business or-
ganization as well as the technique of filmslide pho-
tography is still in the experimental stage.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=123 p:102 ]]

A letter from the Bureau International d'Edition,
46 Rue de la Tour, Paris, states that filmslide publication
of books is already being placed upon a business basis.
An enterprise by the name of "Societé des éditions sur
films des bibliotheques de France" has already begun the
publication of certain rare and valuable books in film-
slide form, using the Photoscopie apparatus. Thus film-
slide copying would compete with book publishing.

Comparison of Efficiencies in Reduction.

Although in theory the length of exposure and
manner of developing have a very great importance in de-
termining the limit of reduction, fineness of focus is
the only contribution to a high reduction ratio that can
be made by the camera equipment itself. Experiments
under non-laboratory conditions show that fineness of
focus is in fact the principal limiting factor. The
difference between the result of an exposure made with
the lens stopped to 6.3 and one made with the stop at
18 is very great, but the difference due to varying
lengths of exposure is small. The finest work has been
done by the Kennedy apparatus which, it happens, employs
an unusually accurate focusing device. Therefore it is
permissible to compare the performance of cameras with

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=124 p:103 ]]

reference to the reduction ratio, even though the camera
and lens are only one of the factors governing resolving
power.

Reduction Ratio and Legibility of 7 Point Newsprint
Photographed in Different Makes of filmslide Camera

Q.R.S. ANSCO LEMARE fiLMOGRAPH EASTMAN LAB LEICA EXPERT LEICA AMAT KENNEDY HIGH KENNEDY LOW

A to 15 to 13 to 16 to 18 to 18 to 20
B 8½ 16-18 13-15 16-20 2 22½ 20-23 21-24
C 15 19-21 16-17 21-24 24-28 25-28 22
D 19 22-25 above 17 25-28 25 29-35 32 29-37
E above 25 above 28 above 36 above 37

The Eastman laboratories concluded on theoreti-
cal grounds that the limit of reduction should be about
25 diameters, and made a test of this reduction ratio,
which had D legibility. The superiority of the Leica to
the Lemare may be due to the extra front lens used in the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=125 p:104 ]]

Leica in copying work. The superior work of the Kenne-
dy apparatus may be due to the use of short wave-length
light and exceptional equipment for focusing. Any appa-
ratus on the market (except the Q.R.S.) will take care
of work which requires a reduction of no more than fif-
teen diameters. The highest reduction possible without
material loss of legibility is 24 diameters.

Projection Apparatus

The weak point in the technique of filmslide
reproduction is not the camera but the projector. The
writer has tested the Leica and Spencer Lens Co. projec-
tors, loaned to the Joint Committee for the purpose.
Both these projectors when used with newsprint reduced
18 diameters will delineate clearly only about one-third
or one-half of the material on the screen. If the center
is brought into focus, the text at the edge of the screen
will be blurred; if the text at the edge is brought into
focus, the center will blur. The two projectors, Leica
and Spencer, do not vary significantly in this respect.
The Spencer Delineascope stands a little further from
the screen, and makes a brighter image than the Leica
machine, but has a coarser focal adjustment. The film
holder of the Leica projector can be turned around to

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=126 ]]
105

show either the side or top of the frame uppermost.
No filmslide increased or diminished in legibility
when it was shifted from the one to the other projec-
tor.

Both projectors are hot to the touch, but the
heat does not seem to damage the film. filmslides left
in the projector for fifty-four hours continuously
showed no ill effects, although throughout this time
they were in direct contact with hot glass. The filmo-
graph organization offers a special projector called
the filmoscope, which is said to project the image
evenly over the whole screen, without blurring at the
edges. The writer has not had an opportunity to test
this device. The Kennedy projector is a modification
of the Leica, and the projector used in the Library of
Congress a modification of the Spencer Delineascope.
In each of these projectors the image is sent from the
filmslide at the reader's side to a mirror, and reflected
back to be read by transmitted light through a sheet of
ground glass. The filmograph projector works from a
mounting above the reader's head, and is read by reflect-
ed light direct from the screen. Experiments with diff-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=127 p:106 ]]

ferent kinds of screens show the surprising result
that the legibility of a filmslide is not much affect-
ed by changing the screen. An image reflected from
white blotting paper is a trifle more legible than one
reflected from the ordinary silver screen. But between
the reflection from the silver screen and transmission
through focusing glass there is no significant diffe-
rence.

Direct Reading of filmslides without Projection.

filmslides of normal-size printed or manuscript
pages, which in the case of non-perforated film are reduc-
tions of seven to ten diameters only, are legible by means
of a powerful reading lens. The reading system designed
for this purpose by Dr. Thomas Martin consists of two
elements: a ground glass stand illuminated from below
which transmits light through a film negative laid upon
it, and a lens fixed at the correct focal distance from
the plans of the glass. The reader merely passes the
film over the ground glass. The fiskoscope could be
used for reading positive paper prints of filmslides, but
the principle of the Martin apparatus is preferable (un-
less it should be necessary to transfer the image from
film to paper in order to preserve it), because it saves

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=128 p:107 ]]

one step in production of the copy, and because it sup-
plies sufficient light.

Choice Among Makes of Equipment on the Market.

If filmslide photography is to attain its max-
imum usefulness, there must be a standardization of equip-
ment which will make rolls of filmslide easily interchange-
able among institutions using them. The most important
choice to be made in this connection lies between the use
of the perforated and non-perforated film. Since the non-
perforated film provides 30% more of sensitive surface than
the perforated, the argument would be all in favor of its
use, were it not that the camera on the market does not
seem to do as fine work as would be expected of it.

The best small camera and equipment now on the
market which uses the non-perforated film is undoubtedly the
"E.K.A." Lemare. But it is so far inferior to the Leica
in resolving power that it loses the advantage of the
extra 20% of film. Leica will reproduce 45 centimeters
with "A" legibility on a 25 millimeter breadth of film;
Lemare will give the same legibility to no more than 39
centimeters on its 30 millimeter breadth of film.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=129 p:108 ]]

The Leica equipment is also likely to be more
useful in meeting the wishes of individual scholars.
The camera can be used for all kinds of photography, the
copying service being only a small part of its general
usefulness. The equipment offered by the Leitz company
is so varied that the purchaser will buy only so much of
the outfit as will be necessary to meet his individual
requirements. A sociologist may wish to copy statistical
material and take pictures of housing conditions; an
anthropologist may wish to use the camera outdoors for
scenes of life among the savages and indoors for making
pictures of small ertefacts as well as copies of pages
of materials that are difficult of access. Someone
travelling in Europe may wish to copy a few books, docu-
ments, or newspapers and use the same camera for making
a photographic record of his vacation. The Ansco appara-
tus has no advantage other than its somewhat lower first
cost. If the Library is equipped with a projector for
Lemare films, the owner of a Leica or Ansco will be able
to read his own films on that projector, or borrow the
films of some other Ansco or Leica owner and read them with-
out additional investment. It is to be hoped that someone

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=130 p:109 ]]

will soon put a better projector on the market.

Costs of Operation.

In the cost of film-slide reproduction there
is a factor analagous to that which affects the cost
of photo-lithographic reproduction - the possible vari-
ety in the ways of distributing labor between the schol-
ar and the technical workshop. If the filmslide camera
is being used by an individual to collect material, it
is likely that he will make the exposures himself. The
filmslide copies will be his substitutes for notes. The
cost per foot of film plus the cost of commercially de-
veloping the negative will be 3½¢ per foot or 4¢ to
allow for wastage of film. For this cash expenditure
the scholar will receive 17 Ansco single frame slides,
7 Leica or Q.R.S. double frame slides, or 6 Lemare non-
perforated double frame slides. All of these prices
are less than normal book costs.

The experience of the Library of Congress
shows that when the labor cost of making the exposures
with Lemare equipment at the leisurely rate of 100 per
day is added to the cost of materials, the cost of each
filmslide (Lemare) is 1½¢. The Harvard Library, using
the Leica equipment, has taken sixty to seventy photo-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=131 p:110 ]]

graphs an hour, at which rate of work the cost would
be much lower. Since all photographing and projecting
equipment will yield "A" legibility at a reduction
ratio of 15. the Lemare filmslide will copy a surface
18 x 24 inches, equivalent to a spread of two pages of
a great 12 x 18 inch folio or news-sheet or one page of
a full size newspaper like the New York Times (18 x 22
inches).

The cost of developing the film is quoted at
2¢ a foot by an Ohio company, and at 1¢ a foot by the
Kinogram corporation. The Universal Talking Newsreel
quotes the high price of 5¢ per foot. From these quo-
tations the following maximum and minimum costs can
be calculated.

If labor costs are cut by using larger machines
such as the filmograph or the Kennedy camera, the cost
per exposure will work out as follows:

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=132 p:111 ]]

HIGH COST LOW COST

If one operator 
uses up one 400 foot 
roll of film, 2500 
exposures. 

If one operator
uses two 400 foot
rolls of film,
5000 exposures
(as claimed)

film….$6.00 $12.00

Devel-
oping....8.00 (2¢ per foot) 8.00 (at cost of l¢ per foot)

Labor
at $1.00
per hr…8.OO 8.00

$22.00 $28.00

The price range then runs from 1½¢ to ½¢ per slide: 1½¢ is
the Library of Congress experience, 1¢ ($.0089) is the cost
if the operator uses 400 feet of film a day and the devel-
oping costs 2¢ a foot, ½¢ ($.0056) is the cost if the op-
erator uses 800 feet of film a day and the developing costs
1¢ a foot. If two pages are put on one slide the minimum
price falls to 28¢ per 100 pages. A single copy is thus
made at a price which book printing does not approach till
the edition rises to one thousand. This is a minimum running
cost, which leaves out of account all expenses of indexing
and storing the finished product, and all overhead costs,
alike for capital and administration. The labor cost at

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=133 p:112 ]]

its lowest stands in a higher ratio to the material
cost in filmslide photography than in photostating.

Conclusions on the Copying of Books and Manuscripts.

The technical perfection of the apparatus
now on the market is sufficient to recommend it for
use in copying all printed and manuscript pages of
normal size at costs far below normal book prices. The
only doubt that can be raised in connection with this
use of the method is the question of the durability of
the film.

Since reproductions on this scale are legi-
ble not only by projection, but also by direct optical
systems such as that designed by Dr. Martin, this cop-
ying procedure can be recommended for general adoption
by libraries to take over some of the load now carried
by the photostat. The photographic equipment of a
large library should include filmslide as well as pho-
tostatic apparatus, because of the lower cost of making
filmslides.

Conclusions on the Copying of Newspapers.

The copying of newspapers by the filmslide
process falls just within present technically practica-
ble limits. If the Lemare camera had Leica precision

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=134 p:113 ]]

or the Leica used non-perforated film, there would be
no difficulty even with small apparatus. The filmslide
method is suited to newspaper copying because it devel-
opes its most significant economies in copying very
large pages, or in making a small number of copies of
a large amount of material. Newspaper reproduction
projects which are impossible because of the prohibi-
tive cost of the photostatic or photo-lithographic
process may become possible if the filmslide method is
used.

The newspaper is a special case, moreover,
because of the high perishability of all newspapers
printed through the last fifty years.

The investigation now under way at the Bureau
of Standards has not determined whether it will be
possible to treat mechanical wood pulp paper in such a
way as to preserve it. If no effective method of pre-
serving ordinary newsprint paper is discovered, or if
the method should prove to be too expensive for general
use, the device which present technology would indicate
as most promising as a means of preserving a fifty
tears' legacy of perishable paper is the filmslide copy-
ing method.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=135 p:114 ]]

The filmslide method also enters into the solu-
tion of the newspaper problem as a device for completing
files, for improving the distribution of the most impor-
tant newspaper sets throughout the country, or for saving
storage space in crowded libraries.

A standard for measuring newspaper costs is fur-
nished by the New York Times. The subscription price is
$18.00 per year. A one-year file, in twenty four bound
volumes, on ordinary newsprint paper, is sold for $72, on
rag paper for $170. The number of pages is close to
24,000, or two thousand a month. The element of bindery
cost in the above quotations is about $2.25 per semi-
monthly volume of one thousand pages. Each volume of the
newsprint issue costs $3.00 bound, and each volume of the
rag paper issue costs $7.08. The cost of a thousand pages
of the rag paper file, exclusive of binding, comes to
$4.83.

At a reduction of 20 diameters two pages of the
New York Times will go on one unperforated filmslide.
The cost could not be higher than 1½¢ nor lower than ½¢
per slide, at which level the following comparisons are
in order:

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=136 p:115 ]]

One Hundred Pages of the New York Times

Rag paper edition, bound in cloth……….$ .71

20 diameter reduction on filmslides
at 1½¢ per slide......................... .75

20 diameter reduction on filmslides
at ½¢ per slide.......... . . . . ............ .28

Newsprint paper edition, bound in cloth.... .30

Cost of binding........... . . . . . . . . . ........ .22½

Reduced to one-half size and printed
by cheapest photo-offset method
in editions of:

25 … $5.02
50 … 2.70
100 … 1.45
200 … .82½
 
One Hundred Pages of the Paris Temps

17 diameter reduction on filmslides at 1½¢ . $1.50
17 diameter reduction on filmslides at ½¢ .. .50
Subscription price & binding in France..... .37
Subscription price & binding in U.S. . . . . . .. .71

One Hundred Pages of the London Times

21 diameter reduction on filmslides at 1½¢. $ .75
21 diameter reduction on filmslides at ½¢.. .25
Subscription & binding.......... . . . . . ...... .33 to .36

One Hundred Pages of Recent European
Newspapers in Back files from Book
Dealers . . . .  .20 to .40

These latter quotations require further explanation.

The Paris Temps prints 2300 pages per year,
each page 20 x 27 inches, slightly larger than the New
York Times. A reduction of 23 diameters would be nec-
essary to put two pages on a filmslide, but 17 diame-
ters would suffice to put one page on a slide. Since

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=137 p:116 ]]

the 23 diameter reduction is close to the top limit
of B legibility, the cost of putting a year's file
on filmslides should be more safely reckoned at one
slide per page, or $12.82 to $34.50 as maximum and
minimum charges. The subscription price in France is
$4.05, in America $12.00 per year. Binding costs
would be $4.50 per year. The total price of buying
and binding a year's issue is therefore $8.55 or
$16.50.

The London Times prints 10,000 pages, each
18½ x 24 inches, per year. This would go on filmslides.
at two pages per slide if reduced 21 diameters - a safe
reduction for B legibility. A year of the Times on
filmslides would cost from $28 to $75. This should be
compared with the subscription and binding price: $33.50
at domestic subscription rate plus binding in five vol-
umes, and $36.55 at the foreign subscription rate.

Recent foreign newspaper files are offered on
the market at comparable prices. A Nijhoff catalogue
offers twelve files of five years each, 1914-1919 at an
average price of $35 per year, bound. A file of Action
Irancaise 1900-1928, 28 years in 68 bound volumes is
offered for $400, or $2.40 per volume, $5.72 per year.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=138 ]]
117.

Most of these newspapers are four or six page issues;
some are eight or twelve page issues. The price per
100 pages of the back files of European newspapers,
including binding, ought to be anything from 20¢ to 40¢.

It must be concluded that filmslide copies of
newspapers can be made at costs within the price range
of files of recent papers, or subscriptions to current
ones. At its lowest, the filmslide edition of a newspaper
may cost but little more than the price of binding the
original edition. If land and building costs are so high
that the expense of providing storage for a newspaper
file exceeds the acquisition cost, it might pay a library
to transfer its bulkier newspaper holdings to filmslides
for the sake of space economy alone, although it must be
recognized that film storage is itself a special problem
in which special fire-proofing and temperature conditions
must be met.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=139 p:118]]

V. Note-Taking Equipment.

The research problem as it comes to the schol-
ar's desk usually calls for a technique of note taking.
The reproduction of research materials in book or film-
slide form ends the work of library institutions. But
the scholar's work begins where that of the library ends.
And here techniques of reproduction ought to be keyed in
with the problem of organizing note systems rather than
the problem of preserving materials.

An illustration of the labor saving possibilities
of new techniques is furnished in the preparation of the
Early Modern English Dictionary under the direction of Pro-
fessor C.C. Fries of Michigan. The same method is to be
used in the New Dictionary of Middle English, under the
direction of Professor Samuel Moore. In dictionary work
the older technique is to look through the literature of
the period for the uses of words, copy out the context of
each use, and file under the word. In preparing these
dictionaries the philologist can now receive from the photo-
lithographer a hundred or more specially made copies of
every book in the form of loose leaves, printed on one
side, and reduced to a standard note size. He need only
underline on a page the words he wishes to take note of;

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=140 p:119 ]]

a typist can then file a separate copy of the whole page
under each of the words so underlined.

The filmslide method in note-taking is best
adapted to the copying of long documents when no rearrange-
ment of the parts is in view, and when it is not necessary
to bring together on the same table at the same time a
number of items to be compared or combined. Mr. Kennedy
has been working out a cheap way of mounting and filing
each filmslide separately upon a small card. This device
would permit the easy grouping and arranging of filmslide
notes, but would not avoid the objection that in using
filmslides the reader can have only one document at a
time before his eyes.

When filmslides are used as a part of a note
system, it may often be found preferable to read them by
means of the Martin magnifying glass and ground glass
equipment rather than by projection. This method is
satisfactory when used in reading filmslides of normal
size books or manuscript pages, although it cannot be
applied to the higher reductions necessary in newspaper
reproduction.

The advertising literature for the Lemare and
Leica apparatus both assume that the user will print

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=141 p:120 ]]

enlarged copies of his filmslides for use as notes.

If this is done, the filmslide is a mere intermediate
step in the production of the note, and the economies
of the method as against the use of the photostat are
largely labor economies arising from the fact that the
scholar makes the exposures himself. If there were on
the market a cheap, portable copying apparatus which
would copy directly on photostat paper of a convenient
note paper size, so arranged that the exposed sheets
would stack themselves in a light-proof box which
could be detached and taken to a photographer's shop
for developing, one step in the photographic note-
taking process would be saved. Notes 5 x 8 inches
would cost less than 2 cents each. Pages of all
sizes would be copied on the same size of note paper,
to file in with notes taken in typescript or manu-
script. The variation in legibility would be more than
compensated for by the convenience of the standardized
size.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=142 p:121 ]]

VI. Recommendations.

1. Library Installation of filmslide Apparatus.

As a substitute for the photostat in copying
books and manuscripts of normal size, present filmslide
apparatus for photography, projection or direct reading
is satisfactory. But it is a misfortune that the came-
ra with the best lens equipment uses the perforated
film, and the only small camera using non-perforated
film has the poorer lens equipment.

It is recommended that the conclusions here
reached in the comparison of these two cameras, the
Leica and Lemare, be scientifically checked, possibly
at the Bureau of Standards.

If the conclusions as to resolving power
here stated are confirmed, - that Leica will give as
fine results at 18 diameters as Lemare at 13, - then
libraries should be urged to install Leica apparatus;
if it is established that the observed differences in
fineness are due to differences in the process of de-
veloping and fixing the film or to differences in ex-
posure time or to some other factor not inherent in the
apparatus, then libraries should be urged to install
Lemare equipment.

In either case, it is very important that all

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=143 p:122 ]]

libraries should install filmslide photography and
projection apparatus as a supplement to the photostat
equipment, and as a device for taking some of the
load off of their lending and borrowing activities.

The cost accounting done at the University
of California established the fact that the expense of
borrowing or lending books is $1.59 per volume -- the
cost of borrowing being almost exactly equal to the
cost of lending. The low cost of filmslide copies
would permit the borrowing scholar to purchase copies
of many items he would otherwise borrow. This is es-
pecially true of short articles in periodicals.

The Joint Committee should take this matter
up with the American Library Association.

2. filmslides and the Newspaper Problem.

As 3 means of copying newspapers, present
photographic apparatus is satisfactory, and present
projection apparatus unsatisfactory. Both can be and
should be improved. It is not recommended that the
filmslide method in its present state be promoted in
any large scale project for preserving or completing
files of newspapers in libraries, although it may be
used as a substitute for the photostat in making

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=144 p:123 ]]

pages of newspapers available to scholars who could
not afford to buy photostats. The possibility of orga-
nizing large-scale copying operations should be kept in
mind as a future part of the program of the Joint Com-
mittee, and in the meantime work upon the following
technical problems should be stimulated through the
Bureau of Standards, the National Council of Research,
or some other agency:

A. A study of the durability of acetate film. If
the investigation discloses that film does not possess
a life probability sufficient to justify its use in the
preservation of records, the possibility of using metal
or glass for small scale record photography should be
explored. If the metal or glass medium is to be used,
the photographic and projecting apparatus would have to
be designed to use the "step and repeat" principle, by
which a large number of images could be taken upon a
single plate.

B. A study of resolving power and reduction ratio,
to determine optimum chemical and optical conditions for
copying newsprint, and to make sure that these are met in
the apparatus in hand with minimum labor cost factor.
Such processes as that invented by Dr. Hutchison should

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=145 p:124 ]]

be tested in this connection.

C. A more efficient projector should be designed.
This might be assigned to an optical company, such as
Bausch and Lomb or the Delineascope for commercial
development.

3. Photostat for Note Taking.

An effort should be made to foster the produc-
tion of a portable and inexpensive photostat equipment
for copying pages of various sizes upon a standard note-
size paper. The Joint Committee should communicate with
the Rectigraph and Photostat Companies, and might even
take measures to verify the reporter's opinion that
there would be a commercial demand for such an apparatus.

4. The Organization of the Book Buying Market in
Connection with Book Production.

The most important recommendation suggested by
this survey is the organizing of the buying power of in-
stitutions which are already devoting their funds to
research purposes. These buyers should control produc-
tion. Works which do not finance themselves as commercial
ventures and must depend on subsidies from the authors or
from endowments or revolving funds could finance them-
selves if the buying power were so organized as to elimi-
nate waste. A survey of scholarly book publication

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=146 p:125 ]]

facilities is being made by Mr. John Marshall under
the direction of the American Council of Learned
Societies. Pending the completion of his survey, one
cannot speak with certainty, but it appears to be true
that the limit which now restricts the circulation of
research material among scholars is set by the require-
ments of publishing finance rather than the purchasing
power of libraries. But the high cost of publication is
due to publishing policies which are not only valueless
to the scholar, but even hostile to his purposes.

It is a consequence of specialization in re-
search that some of the most useful kinds of collabora-
tion among scholars require the circulation of only a
limited number of copies of a given publication. But
commercial publication cannot live on the limited
specialists' market. Therefore the publisher influ-
ences the nature of the product in two ways. He con-
stantly brings pressure to bear upon the productive
scholar to urge him to prepare his work for a wider
audience, and he saddles upon every publishing ven-
ture a burden equal to the whole cost of production of
the book in order to pay for promotion in the market.
The social sciences and humanities are not protected

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=147 p:126 ]]

by the nature of their subject matter from this pres-
sure upon their literature, with the result that even
the purchasing power which is wholly within the control
of the specialists in the field - such as the funds
apportioned to a department in a university library -
must be used in part to fill shelves with material that
is less useful than it should be because of the pressure
upon the author to seek a non-scholarly market.

The learned periodical helps to relieve this
pressure, but not sufficiently, as many instances illus-
trate. A request has come in to the Joint Committee for
funds to continue the publication of a calendar of
Brazilian manuscripts. A conference of research men in
American History, held in Chicago in November 1931, rec-
ommended that facilities for the publication of mono-
graphs intermediate in size between books and articles
in periodicals should be provided. A librarian in the
Cleveland Public Library system, a specialist in Slavic
studies, has compiled a bibliography of English trans-
lations of Polish literature, which is probably worth
publishing but which cannot be financed at present
without a subvention from the author. Source materials
on the Paris Peace Conference ought to be circulated

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=148 p:127]]

and studied as intensively as the source materials on
pre-war diplomacy, and testimony taken while the partic-
ipants in the Conference are still living, but such a
venture under present conditions would not pay its way.
All these items illustrate the problems which could
be met by correlating the organization of book purchas-
ing and book production.

The starting point in organizing purchasing
power in a field is the correlation of two quantities:
the amount of material in a given field which the scholars
therein would wish to publish if publication could be
financed, and the purchasing power which the same schol-
ars share through their ability to control the expendi-
ture of a certain quota of the funds of their respective
libraries.

Any field of scholarship which can organize
its purchasing power in one hundred institutions can be-
gin to take command of production, and to secure for it-
self at normal book prices whatever is deemed most use-
ful for its purposes. An organization of three hundred
can free itself from all restrictions, using the addi-
tional purchasing power either to reduce purchasing
prices or secure better typography and legibility at the

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=149 p:128 ]]

same prices.

The economies which make this control possi-
ble eliminate three-fifths to four-fifths of the nor-
mal costs of publication, by getting rid of:

a. Promotion costs. (The promotion of books in
the scholarly market is done free by the journals any-
way.)

b. Risk of carrying unsold stock. Elimination of
this risk permits the use of more durable paper.

c. Editorial costs, except perhaps for a reading
fee which might be paid in some fields to the scholar
to whom a manuscript is assigned for appraisal. The
cost of correcting spelling and punctuation must not
be levied upon the purchaser.

d. A slight saving in library costs of putting the
book on the shelves can be effected by sending cata-
logue cards along with the books.

e. Authors' royalties and publishers’ profits.

In connection with the development of purchas-
ing orgsnization, materials for publication can be class-
ified into two groups.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=150 p:129 ]]

Group 1, in which the interest of the library predomi-
nates. This may be due to their relation to general
library policy, to their wide appeal, their bibliograph-
ic importance, or their relation to some particular pol-
icy of a special library. Examples: Newspaper reproduc-
tions in a general library; reproductions of state statutes
and law reports in a law library; fac-simile reproductions
of books of general interest which are out of print.

Organization of purchasing power for this class of
material should begin with the libraries. The same kind
of organization which would solve production problems
would also be useful in organizing the completion and
distribution of serial and government document sets.

Group II, in which the interest of the specialized re-
search scholar predominates. The same men are inte-
rested in securing the publication of their own work
and in acquiring the work of others for their own libra-
ries. Examples: Monographic work; biographical source
material concerning persons whose importance lies solely
in one field.

Organization of purchasing power for this class of
material might well begin with the learned societies and

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=151 p:130 ]]

their committees. some or them have publication com-
mittees which could be asked to study a broader pro-
gram based on the possibility of self-financed publication.
Within the A.C.L.S. there are a number of subject matter
committees which might be willing to extend their func-
tions.

Since the practices of institutions vary as to
the responsibilities of the librarian and the research
personnel for the initiative in acquisition, projects
which would be primarily matters of Department discre-
tion in one university might be primarily a matter of
Library responsibility in another. Certainly every
step taken will require complete accord between research
personnel and library. There are some projects which
would be too narrow in scope to be regarded as matters
of general library policy, and yet too broad to fall
within any one research field -- for example, the pub-
lication of Paris Peace Conference Documents. But if
both the libraries and the scholarly personnel are or-
ganized, every worthy project for which there are over
a hundred purchasers can be published and distributed.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=152 p:131 ]]

Action by the Joint Committee

The object described in the foregoing para-
graphs cannot be attained out of hand, by drawing up
schemes for an organization upon paper and having it
voted upon. Existing institutions must be used wherever
possible, and existing wants should be met, as they
arise, in such a way that the new set-up will be
evolved. The following are specific suggestions.

A. Publishing Projects in Search of Funds.

Since any project will finance itself if there
is an advance subscription to the requisite number of
copies, it is not necessary to organize a field or schol-
arship or a system of libraries in order to work out the
practical application of the plan. If particular ventures
are financed in this way, the system will begin to take
root.

Therefore, in all cases in which the Joint Com-
mittee is asked to pass upon the merits of a publishing
project for which a subvention is requested, the action
taken by the committee should have the larger plan in
view. A procedure might well be:

1) To decide whether there is any existing organiza-
tion in the field which could be asked to send our circu-

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=153 p:132 ]]

lars or use its journals to determine whether it con-
trols the purchasing power needed for publication.
Maximum price and minimum legibility should be speci-
fied. The Survey of the Activities of American Agen-
cies in the Improvement of Materials for Research,
just completed for the Joint Committee should be ana-
lyzed to provide lists of special markets for circulariza-
tion in the appropriate fields.

2) To make sure that estimates of editorial costs
are separated from estimated publishing costs; if the
requested subsidy involves both elements, they should
be separately considered.

3) If circularizing of institutions and individuals,
or the sponsorship of the journals, results in a sub-
scription list of more than one hundred, publication can
be financed out of purchase money; if not, the subvention
can be reconsidered.

Among the existing organizations which might
be consulted under (1) above: the publication committees
or the learned societies, especially those which possess
revolving funds, and the Fac-Simile Text society, or
Edwards Brothers if the publication required is a reprint

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=154 p:133 ]]

of an out-of-print publication. The Fac-Simile Text
Society invites suggestions of titles for publication,
and follows the judgment of its editors in making
decisions. Edwards Brothers proposes to receive from
Libraries requests for a fac-simile copy of any man-
uscript or out-of-print book, to circularize libraries
and produce the fac-simile at a price based upon the
number of subscribers.

B. Funds in Search of Publishing Projects.

The Joint Committee might ask scholarly pub-
lishing agencies, committees, revolving funds, etc. to
define their policies in respect of the following two
principles:

1) That photo-lithographed typescript can be sub-
stituted for printing in small editions.

2) That it is neither necessary nor desirable that
the edition of every work published should reach its
maximum size or receive maximum distribution; it is
more important that the number of copies distributed
should be strategically located, since only the great-
est libraries can specialize in everything.

Any field of scholarship in which there is
established a publishing agency which works on these

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=155 p:134 ]]

principles is in a position to take full advantage of
the new publishing processes if its publication commit-
tee can be put directly in touch with the representatives
of buying power. The Joint Committee should help to
create such a liason in these cases.

An arrangement of this kind is under considera-
tion by the officers of the National Council of Teachers
of English. The terms at present under discussion with
Edwards Brothers involve the use of a small revolving
fund.

C. Library Collaboration in Purchasing.

Mr. Gerould at Princeton has warned the Joint
Committee of the probability that the publication of
the check list of Government Documents will be followed
by a system of rare-book prices among the booksellers
dealing in these documents, and has suggested that li-
brary purchasing policies should be worked out in such
a way as to avoid cut-throat competition in acquisition.
Any organization of library purchasing power which
could deal with government document acquisition policies
could also control publication of fac-similes of important
and widely demanded documents which are out of print, and

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=156 p:135 ]]

could pass from that to the control of publication of
other materials in which general library interest pre-
dominates. This problem of Library organization has
already been discussed by the Joint Committee. Any
action taken in connection with Mr. Gerould's sugges-
tion will carry out part of the recommendation result-
ing from this Survey.

D. Newspaper Reproduction.

An experiment should be made to test the
demand for reproductions of old newspaper files in
editions of over one hundred, reproduced by photo-
lithographv. If one important newspaper were selected,
the project explained in the political science and his-
torical publications, and a broader basis of subscription
sought than could be obtained for photostat copies, it is
oossible that a market would be discovered which would
take care of all fac-simile newspaper publication.

Quite distinct from the problem of reproducing
old newspapers is the problem of preserving current news-
papers which are printed on cheap paper. A Survey of
newspaper acquisition policies has been voted by the
Joint Committee, but no work done upon it. If the Survey

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=157 p:136 ]]

should reveal that any newspapers are now being current-
ly bound by more than one hundred libraries, the librar-
ies should be consulted to determine whether they would
wish to have current files reproduced for them half size
upon durable paper. The New York Times, or any news-
paper which is printed on rag stock, would not require
such reproduction. Arrangements might be made in Europe
for the preparation, at the lower European prices, of
photo-lithographed half-size copies of a few of the great
European dailies. Even if the number of libraries
now binding these newspapers is too small to finance the
reproduction scheme, the number might be increased when
the project defines itself.

E. Model Enterprises to be Sponsored by the Joint Committee.

The Joint Committee might select two enterprises
to be sponsored by it and carried through as a demonstra-
tion of the possibilities of bringing organized purchas-
ing power directly into touch with the problem of pub-
lishing. One such enterprise should lie in a field where
there is pressure by research scholars for more adequate
publishing facilities. This would be a problem in setting
up an organization. The other should lie in a field

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=158 p:137 ]]

where there is a demand that certain material be made
more accessible. It would be a matter of carrying
through a project.

Mr. Marshall's report on the use of revolving
funds should suggest the most strategic place in which
to undertake to organize purchasing power to meet the
demand of research publication, thus creating a sample
organization.

Professor Shotwell, now chairman of the
International Relations Committee of the S.S.R.C., has
in mind the preparation of a documentary history of the Paris
Peace Conference. This might perhaps be used as a sample
project. If a sample organization and a sample project
were successful, the scholarly world would begin to think
in new terms of the problem of publication.

F. Further Study of Legibility and Typography.

There is a direct relationship between legibil-
ity, cost, and size of market necessary to float an enter-
prise. The line of investigation begun under the aus-
pices of the Joint Committee should be continued until
three or four standards of legibility are defined, and
the least costly way of producing each of them determined.

It is suggested that In studies of legibility newsprint be

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=159 p:138 ]]

taken as a standard, and that legibilities be graded as
so many points above or below newsprint legibility.
This standard should be selected because newsprint is
familiar to everyone. Professors Paterson and Tinker
of the University of Minnesota might be asked to continue
the work begun in connection with this survey. This
project may fall more nearly within the province of the
National Research Council. It should be pressed forward
in any case.

V. Distribution of this Survey.

This survey is being printed in an edition of
100 copies, free of charge, by Edwards Brothers of Ann
Arbor. Many of these copies have been promised to per-
sons who have furnised information toward the compilation
of the survey. Other copies should be circulated for
criticism by competent persons in the several fields.

The result of criticism, together with the
result of action by the Joint committee, and any further
information that may come to hand, should be worked up
into a new edition of the Survey which can be used in
carrying out the plans of the Joint Committee by present-
ing its point of view in full. The comparison of Leica

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=160 p:139 ]]

and Lemare apparatus should be completed as recommend-
ed before the survey issues in final form. Even the
work on legibility should be developed further than it
is here carried out, so that sample pages of attested
degrees of legibility, each illustrating the lowest
cost of that particular degree of legibility, should
be incorporated in the report. The new edition of the
Survey could then be circulated among the publication
committees of the learned societies and the libraries.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=161 p:[140] ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=162 p:[141] ]]

A SURVEY OF METHODS OF REPRODUCING
RESEARCH MATERIALS

1. Publishing and Library Costs

The Book:- The point of reference from which the problem of
reproducing research materials must be examined is the system by
which publishers produce books and libraries acquire, arrange and
store them for use. The financial policies of institutions, the stack
room and reading room equipment of libraries, the training of library
personnel as well as the research habits of scholars are standardized
in favor of the use of books of several hundred pages, from six to
eleven inches in height, costing the purchaser from one-half to two
cents a page.

There are other forms in which research material is made
accessible, but in any of these other forms there is created a special
problem of finance or administration. Maps, newspapers, "over-
size" books, manuscript collections, films, collections of pamphlets
and broadsides present special difficulties and cause special expanses
regardless of their acquisition costs.

The normal acquisition costs of books or bound volumes of
periodicals which can be shelved like books is something less than
five dollars a volume. In a public library which purchases a large
amount of popular fiction it falls to $1.30 a volume. The list prices
of books per hundred pages, as given in standard annual book-lists
are as follows:

Average Retail Price of Books, Per One Hundred Pages
General Books Scholarly Books (History)

U.S. ................... $ .53 ______________$1.20
Gt. Britain ...............  .86 ................. 1.20
Germany ................  .75 ________________ 1.04
France ..................  .58 .... --(unbound)__ .34
Italy ..................................... .65

SPECIMEN PAGE PRODUCED BY MIME-O-FORM SERVICE, WASHINGTON, D. C.

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=163 ]]

Research materials that can be presented in book form within this
price range can be fed into the library system of the country without
requiring significant changes in present acquisition policies.

Analysis of Book Cost:---The cost accountancy of book pro-
duction is very complicated, but the essential fact is simple: the
unit cost of producing a book falls rapidly as the size of the edition
increases. This is the crux of the problem that must be faced in
any plan which will seek to follow the requirements of research
rather than the demands of the market in deciding what materials
are to be reduced to book form. When the prospective market and
hence the size of the salable edition is small, the unit cost of the
volumes rises above the level at which normal purchasing policies
by or individuals will float the enterprise or accomplish adequate
distribution. The critical test of book cost, for the purposes of this
survey, is the minimum size of an edition which can be sold within
the normal price range.

Another striking characteristic of book costs is the large per-
centage of the sales price which goes to other objects than actual
printing. A great university press calculates that two fifths of
the cost of a book are strictly manufacturing costs, one fifth is edi-
torial and general overhead, and two fifths selling costs. Another
university press bases its price calculation on the following pro-
cedure:

1. Take the "shop cost" of the book (press-work,
paper, binding, etc.) A maximum and minimum
quotation for an edition of 1000, book of 100
pages, 50,000 words is  $370 $775

2. Add 20% for editorial and overhead costs, which
brings the amount to  $444 $930

3. Subtract from the edition 100 copies for review
and promotion, and divide the total cost by the
salable edition (900 copies in this case) to get unit
cost of each book ......... .. $ .49 $1.03

[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=164 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=165 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=166 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=167 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=168 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=169 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=170 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=171 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=172 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=173 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=174 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=175 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=176 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=177 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=178 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=179 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=180 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=181 ]]
[[ http://babel.hathitrust.org/cgi/pt?id=mdp.39015041315394;view=1up;seq=182 ]]